7.11.2.1. chain

7.11.2.1.1. 2n-10ge2p1x710-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 20 vhost 10 chains 10 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 10R1C | 10VM1T
| ... | eth-l2bd-10ch-20vh-10vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${10}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-10c1n

7.11.2.1.1.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.1.2. 64B-1c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.1.3. 64B-2c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.1.4. 64B-4c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.1.5. 1518B-1c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.1.6. 1518B-2c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.1.7. 1518B-4c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.1.8. 9000B-1c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.1.9. 9000B-2c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.1.10. 9000B-4c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.1.11. IMIX-1c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.1.12. IMIX-2c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.1.13. IMIX-4c-eth-l2bd-10ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.2. 2n-10ge2p1x710-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 20 vhost 10 chains 10 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 10R1C | 10VM2T
| ... | eth-l2bd-10ch-20vh-10vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${10}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-10c1n

7.11.2.1.2.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.2.2. 64B-1c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.2.3. 64B-2c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.2.4. 64B-4c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.2.5. 1518B-1c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.2.6. 1518B-2c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.2.7. 1518B-4c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.2.8. 9000B-1c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.2.9. 9000B-2c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.2.10. 9000B-4c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.2.11. IMIX-1c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.2.12. IMIX-2c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.2.13. IMIX-4c-eth-l2bd-10ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.3. 2n-10ge2p1x710-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 40 vhost 10 chains 20 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 10R2C | 20VM1T
| ... | eth-l2bd-10ch-40vh-20vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${10}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-10c2n

7.11.2.1.3.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.3.2. 64B-1c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.3.3. 64B-2c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.3.4. 64B-4c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.3.5. 1518B-1c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.3.6. 1518B-2c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.3.7. 1518B-4c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.3.8. 9000B-1c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.3.9. 9000B-2c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.3.10. 9000B-4c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.3.11. IMIX-1c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.3.12. IMIX-2c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.3.13. IMIX-4c-eth-l2bd-10ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.4. 2n-10ge2p1x710-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 12 vhost 1 chain 6 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R6C | 6VM1T
| ... | eth-l2bd-1ch-12vh-6vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${6}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c6n

7.11.2.1.4.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.4.2. 64B-1c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.4.3. 64B-2c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.4.4. 64B-4c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.4.5. 1518B-1c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.4.6. 1518B-2c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.4.7. 1518B-4c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.4.8. 9000B-1c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.4.9. 9000B-2c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.4.10. 9000B-4c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.4.11. IMIX-1c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.4.12. IMIX-2c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.4.13. IMIX-4c-eth-l2bd-1ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.5. 2n-10ge2p1x710-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 12 vhost 1 chain 6 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R6C | 6VM2T
| ... | eth-l2bd-1ch-12vh-6vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${6}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c6n

7.11.2.1.5.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.5.2. 64B-1c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.5.3. 64B-2c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.5.4. 64B-4c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.5.5. 1518B-1c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.5.6. 1518B-2c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.5.7. 1518B-4c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.5.8. 9000B-1c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.5.9. 9000B-2c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.5.10. 9000B-4c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.5.11. IMIX-1c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.5.12. IMIX-2c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.5.13. IMIX-4c-eth-l2bd-1ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.6. 2n-10ge2p1x710-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 1 chain 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R8C | 8VM1T
| ... | eth-l2bd-1ch-16vh-8vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${8}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c8n

7.11.2.1.6.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.6.2. 64B-1c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.6.3. 64B-2c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.6.4. 64B-4c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.6.5. 1518B-1c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.6.6. 1518B-2c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.6.7. 1518B-4c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.6.8. 9000B-1c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.6.9. 9000B-2c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.6.10. 9000B-4c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.6.11. IMIX-1c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.6.12. IMIX-2c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.6.13. IMIX-4c-eth-l2bd-1ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.7. 2n-10ge2p1x710-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 1 chain 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R8C | 8VM2T
| ... | eth-l2bd-1ch-16vh-8vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${8}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c8n

7.11.2.1.7.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.7.2. 64B-1c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.7.3. 64B-2c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.7.4. 64B-4c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.7.5. 1518B-1c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.7.6. 1518B-2c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.7.7. 1518B-4c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.7.8. 9000B-1c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.7.9. 9000B-2c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.7.10. 9000B-4c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.7.11. IMIX-1c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.7.12. IMIX-2c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.7.13. IMIX-4c-eth-l2bd-1ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.8. 2n-10ge2p1x710-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 20 vhost 1 chain 10 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R10C | 10VM1T
| ... | eth-l2bd-1ch-20vh-10vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${10}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c10n

7.11.2.1.8.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.8.2. 64B-1c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.8.3. 64B-2c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.8.4. 64B-4c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.8.5. 1518B-1c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.8.6. 1518B-2c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.8.7. 1518B-4c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.8.8. 9000B-1c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.8.9. 9000B-2c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.8.10. 9000B-4c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.8.11. IMIX-1c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.8.12. IMIX-2c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.8.13. IMIX-4c-eth-l2bd-1ch-20vh-10vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.9. 2n-10ge2p1x710-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 20 vhost 1 chain 10 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R10C | 10VM2T
| ... | eth-l2bd-1ch-20vh-10vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${10}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c10n

7.11.2.1.9.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.9.2. 64B-1c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.9.3. 64B-2c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.9.4. 64B-4c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.9.5. 1518B-1c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.9.6. 1518B-2c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.9.7. 1518B-4c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.9.8. 9000B-1c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.9.9. 9000B-2c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.9.10. 9000B-4c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.9.11. IMIX-1c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.9.12. IMIX-2c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.9.13. IMIX-4c-eth-l2bd-1ch-20vh-10vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.10. 2n-10ge2p1x710-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 2 vhost 1 chain 1 VM

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R1C | 1VM1T
| ... | eth-l2bd-1ch-2vh-1vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c1n

7.11.2.1.10.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.10.2. 64B-1c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.10.3. 64B-2c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.10.4. 64B-4c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.10.5. 1518B-1c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.10.6. 1518B-2c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.10.7. 1518B-4c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.10.8. 9000B-1c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.10.9. 9000B-2c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.10.10. 9000B-4c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.10.11. IMIX-1c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.10.12. IMIX-2c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.10.13. IMIX-4c-eth-l2bd-1ch-2vh-1vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.11. 2n-10ge2p1x710-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 2 vhost 1 chain 1 VM

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R1C | 1VM2T
| ... | eth-l2bd-1ch-2vh-1vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c1n

7.11.2.1.11.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.11.2. 64B-1c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.11.3. 64B-2c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.11.4. 64B-4c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.11.5. 1518B-1c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.11.6. 1518B-2c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.11.7. 1518B-4c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.11.8. 9000B-1c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.11.9. 9000B-2c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.11.10. 9000B-4c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.11.11. IMIX-1c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.11.12. IMIX-2c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.11.13. IMIX-4c-eth-l2bd-1ch-2vh-1vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.12. 2n-10ge2p1x710-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 4 vhost 1 chain 2 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R2C | 2VM1T
| ... | eth-l2bd-1ch-4vh-2vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c2n

7.11.2.1.12.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.12.2. 64B-1c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.12.3. 64B-2c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.12.4. 64B-4c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.12.5. 1518B-1c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.12.6. 1518B-2c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.12.7. 1518B-4c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.12.8. 9000B-1c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.12.9. 9000B-2c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.12.10. 9000B-4c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.12.11. IMIX-1c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.12.12. IMIX-2c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.12.13. IMIX-4c-eth-l2bd-1ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.13. 2n-10ge2p1x710-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 4 vhost 1 chain 2 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R2C | 2VM2T
| ... | eth-l2bd-1ch-4vh-2vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c2n

7.11.2.1.13.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.13.2. 64B-1c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.13.3. 64B-2c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.13.4. 64B-4c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.13.5. 1518B-1c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.13.6. 1518B-2c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.13.7. 1518B-4c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.13.8. 9000B-1c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.13.9. 9000B-2c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.13.10. 9000B-4c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.13.11. IMIX-1c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.13.12. IMIX-2c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.13.13. IMIX-4c-eth-l2bd-1ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.14. 2n-10ge2p1x710-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 1 chains 4 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R4C | 4VM1T
| ... | eth-l2bd-1ch-8vh-4vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c4n

7.11.2.1.14.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.14.2. 64B-1c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.14.3. 64B-2c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.14.4. 64B-4c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.14.5. 1518B-1c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.14.6. 1518B-2c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.14.7. 1518B-4c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.14.8. 9000B-1c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.14.9. 9000B-2c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.14.10. 9000B-4c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.14.11. IMIX-1c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.14.12. IMIX-2c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.14.13. IMIX-4c-eth-l2bd-1ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.15. 2n-10ge2p1x710-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 1 chains 4 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 1R4C | 4VM2T
| ... | eth-l2bd-1ch-8vh-4vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-1c4n

7.11.2.1.15.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.15.2. 64B-1c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.15.3. 64B-2c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.15.4. 64B-4c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.15.5. 1518B-1c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.15.6. 1518B-2c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.15.7. 1518B-4c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.15.8. 9000B-1c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.15.9. 9000B-2c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.15.10. 9000B-4c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.15.11. IMIX-1c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.15.12. IMIX-2c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.15.13. IMIX-4c-eth-l2bd-1ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.16. 2n-10ge2p1x710-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 2 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R4C | 8VM1T
| ... | eth-l2bd-2ch-16vh-8vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c4n

7.11.2.1.16.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.16.2. 64B-1c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.16.3. 64B-2c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.16.4. 64B-4c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.16.5. 1518B-1c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.16.6. 1518B-2c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.16.7. 1518B-4c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.16.8. 9000B-1c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.16.9. 9000B-2c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.16.10. 9000B-4c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.16.11. IMIX-1c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.16.12. IMIX-2c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.16.13. IMIX-4c-eth-l2bd-2ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.17. 2n-10ge2p1x710-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 2 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R4C | 8VM2T
| ... | eth-l2bd-2ch-16vh-8vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c4n

7.11.2.1.17.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.17.2. 64B-1c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.17.3. 64B-2c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.17.4. 64B-4c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.17.5. 1518B-1c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.17.6. 1518B-2c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.17.7. 1518B-4c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.17.8. 9000B-1c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.17.9. 9000B-2c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.17.10. 9000B-4c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.17.11. IMIX-1c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.17.12. IMIX-2c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.17.13. IMIX-4c-eth-l2bd-2ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.18. 2n-10ge2p1x710-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 24 vhost 2 chains 12 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R6C | 12VM1T
| ... | eth-l2bd-2ch-24vh-12vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${6}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c6n

7.11.2.1.18.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.18.2. 64B-1c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.18.3. 64B-2c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.18.4. 64B-4c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.18.5. 1518B-1c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.18.6. 1518B-2c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.18.7. 1518B-4c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.18.8. 9000B-1c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.18.9. 9000B-2c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.18.10. 9000B-4c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.18.11. IMIX-1c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.18.12. IMIX-2c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.18.13. IMIX-4c-eth-l2bd-2ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.19. 2n-10ge2p1x710-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 24 vhost 2 chains 12 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R6C | 12VM2T
| ... | eth-l2bd-2ch-24vh-12vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${6}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c6n

7.11.2.1.19.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.19.2. 64B-1c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.19.3. 64B-2c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.19.4. 64B-4c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.19.5. 1518B-1c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.19.6. 1518B-2c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.19.7. 1518B-4c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.19.8. 9000B-1c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.19.9. 9000B-2c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.19.10. 9000B-4c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.19.11. IMIX-1c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.19.12. IMIX-2c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.19.13. IMIX-4c-eth-l2bd-2ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.20. 2n-10ge2p1x710-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 2 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R8C | 16VM1T
| ... | eth-l2bd-2ch-32vh-16vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${8}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c8n

7.11.2.1.20.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.20.2. 64B-1c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.20.3. 64B-2c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.20.4. 64B-4c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.20.5. 1518B-1c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.20.6. 1518B-2c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.20.7. 1518B-4c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.20.8. 9000B-1c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.20.9. 9000B-2c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.20.10. 9000B-4c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.20.11. IMIX-1c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.20.12. IMIX-2c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.20.13. IMIX-4c-eth-l2bd-2ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.21. 2n-10ge2p1x710-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 2 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R8C | 16VM2T
| ... | eth-l2bd-2ch-32vh-16vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${8}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c8n

7.11.2.1.21.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.21.2. 64B-1c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.21.3. 64B-2c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.21.4. 1518B-1c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.21.5. 1518B-2c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.21.6. 9000B-1c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.21.7. 9000B-2c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.21.8. IMIX-1c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.21.9. IMIX-2c-eth-l2bd-2ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.22. 2n-10ge2p1x710-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 40 vhost 2 chains 20 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R10C | 20VM1T
| ... | eth-l2bd-2ch-40vh-20vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${10}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c10n

7.11.2.1.22.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.22.2. 64B-1c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.22.3. 64B-2c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.22.4. 64B-4c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.22.5. 1518B-1c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.22.6. 1518B-2c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.22.7. 1518B-4c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.22.8. 9000B-1c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.22.9. 9000B-2c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.22.10. 9000B-4c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.22.11. IMIX-1c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.22.12. IMIX-2c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.22.13. IMIX-4c-eth-l2bd-2ch-40vh-20vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.23. 2n-10ge2p1x710-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 4 vhost 2 chains 2 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R1C | 2VM1T
| ... | eth-l2bd-2ch-4vh-2vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c1n

7.11.2.1.23.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.23.2. 64B-1c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.23.3. 64B-2c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.23.4. 64B-4c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.23.5. 1518B-1c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.23.6. 1518B-2c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.23.7. 1518B-4c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.23.8. 9000B-1c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.23.9. 9000B-2c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.23.10. 9000B-4c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.23.11. IMIX-1c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.23.12. IMIX-2c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.23.13. IMIX-4c-eth-l2bd-2ch-4vh-2vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.24. 2n-10ge2p1x710-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 4 vhost 2 chains 2 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R1C | 2VM2T
| ... | eth-l2bd-2ch-4vh-2vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c1n

7.11.2.1.24.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.24.2. 64B-1c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.24.3. 64B-2c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.24.4. 64B-4c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.24.5. 1518B-1c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.24.6. 1518B-2c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.24.7. 1518B-4c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.24.8. 9000B-1c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.24.9. 9000B-2c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.24.10. 9000B-4c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.24.11. IMIX-1c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.24.12. IMIX-2c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.24.13. IMIX-4c-eth-l2bd-2ch-4vh-2vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.25. 2n-10ge2p1x710-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 2 chains 4 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R2C | 4VM1T
| ... | eth-l2bd-2ch-8vh-4vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c2n

7.11.2.1.25.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.25.2. 64B-1c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.25.3. 64B-2c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.25.4. 64B-4c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.25.5. 1518B-1c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.25.6. 1518B-2c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.25.7. 1518B-4c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.25.8. 9000B-1c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.25.9. 9000B-2c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.25.10. 9000B-4c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.25.11. IMIX-1c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.25.12. IMIX-2c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.25.13. IMIX-4c-eth-l2bd-2ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.26. 2n-10ge2p1x710-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 2 chains 4 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 2R2C | 4VM2T
| ... | eth-l2bd-2ch-8vh-4vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${2}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-2c2n

7.11.2.1.26.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.26.2. 64B-1c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.26.3. 64B-2c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.26.4. 64B-4c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.26.5. 1518B-1c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.26.6. 1518B-2c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.26.7. 1518B-4c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.26.8. 9000B-1c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.26.9. 9000B-2c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.26.10. 9000B-4c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.26.11. IMIX-1c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.26.12. IMIX-2c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.26.13. IMIX-4c-eth-l2bd-2ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.27. 2n-10ge2p1x710-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 4 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R2C | 8VM1T
| ... | eth-l2bd-4ch-16vh-8vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c2n

7.11.2.1.27.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.27.2. 64B-1c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.27.3. 64B-2c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.27.4. 64B-4c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.27.5. 1518B-1c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.27.6. 1518B-2c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.27.7. 1518B-4c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.27.8. 9000B-1c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.27.9. 9000B-2c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.27.10. 9000B-4c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.27.11. IMIX-1c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.27.12. IMIX-2c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.27.13. IMIX-4c-eth-l2bd-4ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.28. 2n-10ge2p1x710-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 4 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R2C | 8VM2T
| ... | eth-l2bd-4ch-16vh-8vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c2n

7.11.2.1.28.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.28.2. 64B-1c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.28.3. 64B-2c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.28.4. 64B-4c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.28.5. 1518B-1c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.28.6. 1518B-2c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.28.7. 1518B-4c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.28.8. 9000B-1c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.28.9. 9000B-2c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.28.10. 9000B-4c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.28.11. IMIX-1c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.28.12. IMIX-2c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.28.13. IMIX-4c-eth-l2bd-4ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.29. 2n-10ge2p1x710-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 4 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R4C | 16VM1T
| ... | eth-l2bd-4ch-32vh-16vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c4n

7.11.2.1.29.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.29.2. 64B-1c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.29.3. 64B-2c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.29.4. 64B-4c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.29.5. 1518B-1c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.29.6. 1518B-2c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.29.7. 1518B-4c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.29.8. 9000B-1c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.29.9. 9000B-2c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.29.10. 9000B-4c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.29.11. IMIX-1c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.29.12. IMIX-2c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.29.13. IMIX-4c-eth-l2bd-4ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.30. 2n-10ge2p1x710-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 4 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R4C | 16VM2T
| ... | eth-l2bd-4ch-32vh-16vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c4n

7.11.2.1.30.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.30.2. 64B-1c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.30.3. 64B-2c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.30.4. 1518B-1c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.30.5. 1518B-2c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.30.6. 9000B-1c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.30.7. 9000B-2c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.30.8. IMIX-1c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.30.9. IMIX-2c-eth-l2bd-4ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.31. 2n-10ge2p1x710-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 48 vhost 4 chains 24 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R6C | 24VM1T
| ... | eth-l2bd-4ch-48vh-24vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${6}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c6n

7.11.2.1.31.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.31.2. 64B-1c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.31.3. 64B-2c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.31.4. 1518B-1c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.31.5. 1518B-2c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.31.6. 9000B-1c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.31.7. 9000B-2c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.31.8. IMIX-1c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.31.9. IMIX-2c-eth-l2bd-4ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.32. 2n-10ge2p1x710-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 4 chains 4 VM

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R1C | 4VM1T
| ... | eth-l2bd-4ch-8vh-4vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c1n

7.11.2.1.32.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.32.2. 64B-1c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.32.3. 64B-2c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.32.4. 64B-4c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.32.5. 1518B-1c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.32.6. 1518B-2c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.32.7. 1518B-4c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.32.8. 9000B-1c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.32.9. 9000B-2c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.32.10. 9000B-4c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.32.11. IMIX-1c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.32.12. IMIX-2c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.32.13. IMIX-4c-eth-l2bd-4ch-8vh-4vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.33. 2n-10ge2p1x710-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 8 vhost 4 chains 4 VM

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 4R1C | 4VM2T
| ... | eth-l2bd-4ch-8vh-4vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${4}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-4c1n

7.11.2.1.33.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.33.2. 64B-1c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.33.3. 64B-2c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.33.4. 64B-4c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.33.5. 1518B-1c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.33.6. 1518B-2c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.33.7. 1518B-4c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.33.8. 9000B-1c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.33.9. 9000B-2c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.33.10. 9000B-4c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.33.11. IMIX-1c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.33.12. IMIX-2c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.33.13. IMIX-4c-eth-l2bd-4ch-8vh-4vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.34. 2n-10ge2p1x710-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 12 vhost 6 chains 6 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 6R1C | 6VM1T
| ... | eth-l2bd-6ch-12vh-6vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${6}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-6c1n

7.11.2.1.34.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.34.2. 64B-1c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.34.3. 64B-2c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.34.4. 64B-4c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.34.5. 1518B-1c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.34.6. 1518B-2c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.34.7. 1518B-4c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.34.8. 9000B-1c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.34.9. 9000B-2c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.34.10. 9000B-4c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.34.11. IMIX-1c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.34.12. IMIX-2c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.34.13. IMIX-4c-eth-l2bd-6ch-12vh-6vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.35. 2n-10ge2p1x710-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 12 vhost 6 chains 6 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 6R1C | 6VM2T
| ... | eth-l2bd-6ch-12vh-6vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${6}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-6c1n

7.11.2.1.35.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.35.2. 64B-1c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.35.3. 64B-2c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.35.4. 64B-4c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.35.5. 1518B-1c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.35.6. 1518B-2c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.35.7. 1518B-4c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.35.8. 9000B-1c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.35.9. 9000B-2c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.35.10. 9000B-4c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.35.11. IMIX-1c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.35.12. IMIX-2c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.35.13. IMIX-4c-eth-l2bd-6ch-12vh-6vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.36. 2n-10ge2p1x710-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 24 vhost 6 chains 12 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 6R2C | 12VM1T
| ... | eth-l2bd-6ch-24vh-12vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${6}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-6c2n

7.11.2.1.36.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.36.2. 64B-1c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.36.3. 64B-2c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.36.4. 64B-4c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.36.5. 1518B-1c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.36.6. 1518B-2c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.36.7. 1518B-4c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.36.8. 9000B-1c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.36.9. 9000B-2c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.36.10. 9000B-4c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.36.11. IMIX-1c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.36.12. IMIX-2c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.36.13. IMIX-4c-eth-l2bd-6ch-24vh-12vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.37. 2n-10ge2p1x710-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 24 vhost 6 chains 12 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 6R2C | 12VM2T
| ... | eth-l2bd-6ch-24vh-12vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${6}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-6c2n

7.11.2.1.37.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.37.2. 64B-1c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.37.3. 64B-2c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.37.4. 64B-4c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.37.5. 1518B-1c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.37.6. 1518B-2c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.37.7. 1518B-4c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.37.8. 9000B-1c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.37.9. 9000B-2c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.37.10. 9000B-4c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.37.11. IMIX-1c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.37.12. IMIX-2c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.37.13. IMIX-4c-eth-l2bd-6ch-24vh-12vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.38. 2n-10ge2p1x710-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 48 vhost 6 chains 24 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 6R4C | 24VM1T
| ... | eth-l2bd-6ch-48vh-24vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${6}
| ${nf_nodes}= | ${4}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-6c4n

7.11.2.1.38.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.38.2. 64B-1c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.38.3. 64B-2c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.38.4. 1518B-1c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.38.5. 1518B-2c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.38.6. 9000B-1c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.38.7. 9000B-2c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.38.8. IMIX-1c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.38.9. IMIX-2c-eth-l2bd-6ch-48vh-24vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.39. 2n-10ge2p1x710-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 8 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 8R1C | 8VM1T
| ... | eth-l2bd-8ch-16vh-8vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${8}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-8c1n

7.11.2.1.39.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.39.2. 64B-1c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.39.3. 64B-2c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.39.4. 64B-4c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.39.5. 1518B-1c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.39.6. 1518B-2c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.39.7. 1518B-4c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.39.8. 9000B-1c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.39.9. 9000B-2c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.39.10. 9000B-4c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.39.11. IMIX-1c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.39.12. IMIX-2c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.39.13. IMIX-4c-eth-l2bd-8ch-16vh-8vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.40. 2n-10ge2p1x710-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 16 vhost 8 chains 8 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 8R1C | 8VM2T
| ... | eth-l2bd-8ch-16vh-8vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${8}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-8c1n

7.11.2.1.40.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.40.2. 64B-1c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.40.3. 64B-2c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.40.4. 64B-4c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.40.5. 1518B-1c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.40.6. 1518B-2c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.40.7. 1518B-4c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.40.8. 9000B-1c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.40.9. 9000B-2c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.40.10. 9000B-4c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.40.11. IMIX-1c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.40.12. IMIX-2c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.40.13. IMIX-4c-eth-l2bd-8ch-16vh-8vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.41. 2n-10ge2p1x710-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 8 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 8R2C | 16VM1T
| ... | eth-l2bd-8ch-32vh-16vm1t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${2}
| ${nf_dtc}= | ${0.5}
| ${nf_chains}= | ${8}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-8c2n

7.11.2.1.41.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.41.2. 64B-1c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.41.3. 64B-2c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.41.4. 64B-4c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${4}

7.11.2.1.41.5. 1518B-1c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.41.6. 1518B-2c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.41.7. 1518B-4c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${4}

7.11.2.1.41.8. 9000B-1c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.41.9. 9000B-2c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.41.10. 9000B-4c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${4}

7.11.2.1.41.11. IMIX-1c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.41.12. IMIX-2c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}

7.11.2.1.41.13. IMIX-4c-eth-l2bd-8ch-32vh-16vm1t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${4}

7.11.2.1.42. 2n-10ge2p1x710-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr suite

RFC2544: Pkt throughput L2BD test cases with 32 vhost 8 chains 16 VMs

  • [Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.

  • [Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.

  • [Cfg] DUT configuration: DUT1 is configured with L2 bridge- domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP IPv4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1 is tested with ${nic_name}.

  • [Ver] TG verification: TG finds and reports throughput NDR (Non Drop Rate) with zero packet loss tolerance and throughput PDR (Partial Drop Rate) with non-zero packet loss tolerance (LT) expressed in percentage of packets transmitted. NDR and PDR are discovered for different Ethernet L2 frame sizes using MLRsearch library.

Test packets are generated by TG on links to DUTs. TG traffic profile contains two L3 flow-groups (flow-group per direction, 254 flows per flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61 and static payload. MAC addresses are matching MAC addresses of NFs nodes interfaces.

  • [Ref] Applicable standard specifications: RFC2544.

*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 2_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | VM | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | VHOST_1024 | NF_DENSITY | NF_VPPIP4 | CHAIN | 8R2C | 16VM2T
| ... | eth-l2bd-8ch-32vh-16vm2t-vppip4
|
| Suite Setup | Setup suite topology interfaces | performance
| Suite Teardown | Tear down suite | performance
| Test Setup | Setup test | performance
| Test Teardown | Tear down test | performance | vhost
|
| Test Template | Local Template
|
| 
*** Variables ***
| @{plugins_to_enable}= | dpdk_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 2
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${8}
| ${nf_nodes}= | ${2}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n3n-ethip4-ip4src254-8c2n

7.11.2.1.42.1. Local Template

  • [Cfg] DUT runs L2BD switching config. Each DUT uses ${phy_cores} physical core(s) for worker threads.

  • [Ver] Measure NDR and PDR values using MLRsearch algorithm.

Arguments: - frame_size - Framesize in Bytes in integer or string (IMIX_v4_1). Type: integer, string - phy_cores - Number of physical cores. Type: integer - rxq - Number of RX queues, default value: ${None}. Type: integer


Set Test Variable  \${frame_size}
Given Set Max Rate And Jumbo
  And Add worker threads to all DUTs  ${phy_cores}  ${rxq}
  And Pre-initialize layer driver  ${nic_driver}
  And Apply startup configuration on all VPP DUTs
 When Initialize layer driver  ${nic_driver}
  And Initialize layer interface  count=${nf_chains}
  And Initialize L2 bridge domains for multiple chains with Vhost-User  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}
  And Configure chains of NFs connected via vhost-user  nf_chains=${nf_chains}  nf_nodes=${nf_nodes}  jumbo=${jumbo}  use_tuned_cfs=${False}  auto_scale=${False}  vnf=vppip4_2vhostvr1024
 Then Find NDR and PDR intervals using optimized search

7.11.2.1.42.2. 64B-1c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${1}

7.11.2.1.42.3. 64B-2c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${64}  phy_cores=${2}

7.11.2.1.42.4. 1518B-1c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${1}

7.11.2.1.42.5. 1518B-2c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${1518}  phy_cores=${2}

7.11.2.1.42.6. 9000B-1c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${1}

7.11.2.1.42.7. 9000B-2c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=${9000}  phy_cores=${2}

7.11.2.1.42.8. IMIX-1c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${1}

7.11.2.1.42.9. IMIX-2c-eth-l2bd-8ch-32vh-16vm2t-vppip4-ndrpdr


frame_size=IMIX_v4_1  phy_cores=${2}