7.13. vm_vhost¶
7.13.1. 10ge2p1x710-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 bridge-domain with MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_1L | VM_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_plugin.so | lacp_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
| ${bd_id1}= | 1
| ${bd_id2}= | 2
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.1.1. 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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
And Initialize L2 bridge domains with Vhost-User and VLAN with VPP link bonding in a 3-node circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.1.2. 64B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.1.3. 64B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.1.4. 64B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.1.5. 1518B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.1.6. 1518B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.1.7. 1518B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.1.8. 9000B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.1.9. 9000B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.1.10. 9000B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.1.11. IMIX-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.1.12. IMIX-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.1.13. IMIX-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.2. 10ge2p1x710-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 bridge-domain with MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_1L | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_plugin.so | lacp_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
| ${bd_id1}= | 1
| ${bd_id2}= | 2
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.2.1. 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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
And Initialize L2 bridge domains with Vhost-User and VLAN with VPP link bonding in a 3-node circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.2.2. 64B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.2.3. 64B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.2.4. 64B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.2.5. 1518B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.2.6. 1518B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.2.7. 1518B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.2.8. 9000B-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.2.9. 9000B-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.2.10. 9000B-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.2.11. IMIX-1c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.2.12. IMIX-2c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.2.13. IMIX-4c-1lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.3. 10ge2p1x710-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_1L | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_plugin.so | lacp_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.3.1. Local Template¶
[Cfg] DUT runs L2XC 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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
And Initialize L2 xconnect with Vhost-User and VLAN with VPP link bonding in 3-node circular topology ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.3.2. 64B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.3.3. 64B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.3.4. 64B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.3.5. 1518B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.3.6. 1518B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.3.7. 1518B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.3.8. 9000B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.3.9. 9000B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.3.10. 9000B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.3.11. IMIX-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.3.12. IMIX-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.3.13. IMIX-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.4. 10ge2p1x710-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_1L | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_plugin.so | lacp_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.4.1. Local Template¶
[Cfg] DUT runs L2XC 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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
And Initialize L2 xconnect with Vhost-User and VLAN with VPP link bonding in 3-node circular topology ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.4.2. 64B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.4.3. 64B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.4.4. 64B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.4.5. 1518B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.4.6. 1518B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.4.7. 1518B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.4.8. 9000B-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.4.9. 9000B-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.4.10. 9000B-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.4.11. IMIX-1c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.4.12. IMIX-2c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.4.13. IMIX-4c-1lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.5. 10ge2p1x710-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1=DUT2-TG 3-node circular topology with single links between TG and DUT nodes and double link between DUT nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 bridge-domain with MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_DOUBLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_2L | VM_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_plugin.so | lacp_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 4
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
| ${bd_id1}= | 1
| ${bd_id2}= | 2
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.5.1. 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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 L2 bridge domains with Vhost-User and VLAN with VPP link bonding in a 3-node circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.5.2. 64B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.5.3. 64B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.5.4. 64B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.5.5. 1518B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.5.6. 1518B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.5.7. 1518B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.5.8. 9000B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.5.9. 9000B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.5.10. 9000B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.5.11. IMIX-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.5.12. IMIX-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.5.13. IMIX-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.6. 10ge2p1x710-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1=DUT2-TG 3-node circular topology with single links between TG and DUT nodes and double link between DUT nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 bridge-domain with MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_DOUBLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_2L | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_plugin.so | lacp_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 4
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
| ${bd_id1}= | 1
| ${bd_id2}= | 2
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.6.1. 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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 L2 bridge domains with Vhost-User and VLAN with VPP link bonding in a 3-node circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.6.2. 64B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.6.3. 64B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.6.4. 64B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.6.5. 1518B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.6.6. 1518B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.6.7. 1518B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.6.8. 9000B-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.6.9. 9000B-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.6.10. 9000B-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.6.11. IMIX-1c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.6.12. IMIX-2c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.6.13. IMIX-4c-2lbvpplacp-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.7. 10ge2p1x710-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1=DUT2-TG 3-node circular topology with single links between TG and DUT nodes and double link between DUT nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_DOUBLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_2L | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_plugin.so | lacp_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 4
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.7.1. Local Template¶
[Cfg] DUT runs L2XC 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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 L2 xconnect with Vhost-User and VLAN with VPP link bonding in 3-node circular topology ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.7.2. 64B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.7.3. 64B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.7.4. 64B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.7.5. 1518B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.7.6. 1518B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.7.7. 1518B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.7.8. 9000B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.7.9. 9000B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.7.10. 9000B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.7.11. IMIX-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.7.12. IMIX-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.7.13. IMIX-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.8. 10ge2p1x710-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost and vpp link bonding
[Top] Network Topologies: TG-DUT1=DUT2-TG 3-node circular topology with single links between TG and DUT nodes and double link between DUT nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with VPP link bonding (mode LACP, transmit policy l34) on link between DUT1 and DUT2 and L2 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_DOUBLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST |1VM
| ... | VHOST_1024 | LBOND | LBOND_VPP | LBOND_MODE_LACP | LBOND_LB_L34
| ... | LBOND_2L | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | 2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_plugin.so | lacp_plugin.so
| ${crypto_type}= | ${None}
| ${nic_name}= | Intel-X710
| ${nic_driver}= | vfio-pci
| ${nic_rxq_size}= | 0
| ${nic_txq_size}= | 0
| ${nic_pfs}= | 4
| ${nic_vfs}= | 0
| ${osi_layer}= | L2
| ${overhead}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Link bonding config
| ${bond_mode}= | lacp
| ${lb_mode}= | l34
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.8.1. Local Template¶
[Cfg] DUT runs L2XC 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}
Set Test Variable ${vlan_strip_off} ${TRUE}
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 L2 xconnect with Vhost-User and VLAN with VPP link bonding in 3-node circular topology ${subid} ${tag_rewrite} ${bond_mode} ${lb_mode}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.8.2. 64B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.8.3. 64B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.8.4. 64B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.8.5. 1518B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.8.6. 1518B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.8.7. 1518B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.8.8. 9000B-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.8.9. 9000B-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.8.10. 9000B-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.8.11. IMIX-1c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.8.12. IMIX-2c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.8.13. IMIX-4c-2lbvpplacp-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.9. 10ge2p1x710-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD with vhost and IEEE 802.1Q test cases
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. IEEE 802.1Q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.9.1. Local Template¶
[Cfg] Each DUT runs L2BD switching with VLAN and 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
And Initialize L2 bridge domains with Vhost-User and VLAN in circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.9.2. 64B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.9.3. 64B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.9.4. 64B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.9.5. 1518B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.9.6. 1518B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.9.7. 1518B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.9.8. 9000B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.9.9. 9000B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.9.10. 9000B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.9.11. IMIX-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.9.12. IMIX-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.9.13. IMIX-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.10. 10ge2p1x710-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD with vhost abd IEEE 802.1Q test cases
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. IEEE 802.1Q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.10.1. Local Template¶
[Cfg] Each DUT runs L2BD switching with VLAN and 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
And Initialize L2 bridge domains with Vhost-User and VLAN in circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.10.2. 64B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.10.3. 64B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.10.4. 64B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.10.5. 1518B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.10.6. 1518B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.10.7. 1518B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.10.8. 9000B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.10.9. 9000B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.10.10. 9000B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.10.11. IMIX-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.10.12. IMIX-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.10.13. IMIX-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.11. 10ge2p1x710-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 cross-connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.11.1. Local Template¶
[Cfg] DUT runs L2XC 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
And Initialize L2 xconnect with Vhost-User and VLAN in circular topology ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.11.2. 64B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.11.3. 64B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.11.4. 64B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.11.5. 1518B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.11.6. 1518B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.11.7. 1518B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.11.8. 9000B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.11.9. 9000B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.11.10. 9000B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.11.11. IMIX-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.11.12. IMIX-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.11.13. IMIX-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.12. 10ge2p1x710-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect. 802.1q tagging is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 cross-connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.12.1. Local Template¶
[Cfg] DUT runs L2XC 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
And Initialize L2 xconnect with Vhost-User and VLAN in circular topology ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.12.2. 64B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.12.3. 64B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.12.4. 64B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.12.5. 1518B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.12.6. 1518B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.12.7. 1518B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.12.8. 9000B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.12.9. 9000B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.12.10. 9000B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.12.11. IMIX-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.12.12. IMIX-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.12.13. IMIX-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.13. 10ge2p1x710-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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, 253 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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-3n-ethip4-ip4src2547.13.13.1. 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=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.13.2. 64B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.13.3. 64B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.13.4. 64B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.13.5. 1518B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.13.6. 1518B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.13.7. 1518B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.13.8. 9000B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.13.9. 9000B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.13.10. 9000B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.13.11. IMIX-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.13.12. IMIX-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.13.13. IMIX-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.14. 10ge2p1x710-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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, 253 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2BDMACLRN | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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-3n-ethip4-ip4src2547.13.14.1. 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=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.14.2. 64B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.14.3. 64B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.14.4. 64B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.14.5. 1518B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.14.6. 1518B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.14.7. 1518B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.14.8. 9000B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.14.9. 9000B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.14.10. 9000B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.14.11. IMIX-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.14.12. IMIX-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.14.13. IMIX-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.15. 10ge2p1x710-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 cross-connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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-3n-ethip4-ip4src2547.13.15.1. Local Template¶
[Cfg] DUT runs L2XC 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 ${nic_driver}
And Initialize layer interface
And Initialize L2 xconnect with Vhost-User nf_nodes=${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=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.15.2. 64B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.15.3. 64B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.15.4. 64B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.15.5. 1518B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.15.6. 1518B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.15.7. 1518B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.15.8. 9000B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.15.9. 9000B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.15.10. 9000B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.15.11. IMIX-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.15.12. IMIX-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.15.13. IMIX-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.16. 10ge2p1x710-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 cross-connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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-3n-ethip4-ip4src2547.13.16.1. Local Template¶
[Cfg] DUT runs L2XC 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 ${nic_driver}
And Initialize layer interface
And Initialize L2 xconnect with Vhost-User nf_nodes=${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=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.16.2. 64B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.16.3. 64B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.16.4. 64B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.16.5. 1518B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.16.6. 1518B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.16.7. 1518B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.16.8. 9000B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.16.9. 9000B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.16.10. 9000B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.16.11. IMIX-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.16.12. IMIX-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.16.13. IMIX-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.17. 10ge2p1x710-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput IPv4 test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for IPv4 routing.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with IPv4. routing and two static IPv4 /24 route entries. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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, 253 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | IP4FWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4-ip4base-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | L3
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2537.13.17.1. Local Template¶
[Cfg] DUT runs IPv4 routing 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 ${nic_driver}
And Initialize layer interface
And Initialize IPv4 forwarding with vhost in 3-node circular topology nf_nodes=${nf_nodes} testpmd_mac=${TRUE}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_mac
Then Find NDR and PDR intervals using optimized search
7.13.17.2. 64B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.17.3. 64B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.17.4. 64B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.17.5. 1518B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.17.6. 1518B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.17.7. 1518B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.17.8. 9000B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.17.9. 9000B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.17.10. 9000B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.17.11. IMIX-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.17.12. IMIX-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.17.13. IMIX-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.18. 10ge2p1x710-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr suite¶
RFC2544: Pkt throughput IPv4 test cases with vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for IPv4 routing.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with IPv4. routing and two static IPv4 /24 route entries. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP ip4 interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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, 253 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC2544.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | ETH | IP4FWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPIP4 | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4-ip4base-eth-2vhostvr1024-1vm-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 | perfmon_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}= | L3
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2537.13.18.1. Local Template¶
[Cfg] DUT runs IPv4 routing 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 ${nic_driver}
And Initialize layer interface
And Initialize IPv4 forwarding with vhost in 3-node circular topology nf_nodes=${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=${True} vnf=vppip4noarp_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.18.2. 64B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.18.3. 64B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.18.4. 64B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.18.5. 1518B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.18.6. 1518B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.18.7. 1518B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.18.8. 9000B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.18.9. 9000B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.18.10. 9000B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.18.11. IMIX-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.18.12. IMIX-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.18.13. IMIX-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.19. 10ge2p1x710-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with VXLANoIPv4. and vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. Eth-IPv4-VXLAN-Eth-IPv4 is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC7348.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | L2BDMACLRN | ENCAP | VXLAN | L2OVRLAY | IP4UNRLAY
| ... | VHOST | 1VM | VHOST_1024 | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | ${50}
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.19.1. 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
And Initialize L2 bridge domains with Vhost-User and VXLANoIPv4 in 3-node circular topology ${bd_id1} ${bd_id2}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_chains} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.19.2. 64B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.19.3. 64B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.19.4. 64B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.19.5. 1518B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.19.6. 1518B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.19.7. 1518B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.19.8. 9000B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.19.9. 9000B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.19.10. 9000B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.19.11. IMIX-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.19.12. IMIX-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.19.13. IMIX-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.20. 10ge2p1x710-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with VXLANoIPv4. and vhost
[Top] Network Topologies: TG-DUT1-DUT2-TG 3-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 switching of IPv4. Eth-IPv4-VXLAN-Eth-IPv4 is applied on link between DUT1 and DUT2.
[Cfg] DUT configuration: DUT1 and DUT2 are configured with L2 bridge-domain and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting vhost-user interfaces, rxd/txd=1024. DUT1/DUT2 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 the TG node interfaces.
[Ref] Applicable standard specifications: RFC7348.
*** Settings ***
| Resource | resources/libraries/robot/shared/default.robot
|
| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | NDRPDR
| ... | NIC_Intel-X710 | L2BDMACLRN | ENCAP | VXLAN | L2OVRLAY | IP4UNRLAY
| ... | VHOST | 1VM | VHOST_1024 | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${50}
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-3n-ethip4-ip4src2547.13.20.1. 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
And Initialize L2 bridge domains with Vhost-User and VXLANoIPv4 in 3-node circular topology ${bd_id1} ${bd_id2}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_chains} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.20.2. 64B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.20.3. 64B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.20.4. 64B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.20.5. 1518B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.20.6. 1518B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.20.7. 1518B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.20.8. 9000B-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.20.9. 9000B-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.20.10. 9000B-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.20.11. IMIX-1c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.20.12. IMIX-2c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.20.13. IMIX-4c-ethip4vxlan-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.21. 2n1l-10ge2p1x710-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD with vhost and IEEE 802.1Q test cases
[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. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[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 testpmd 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 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.21.1. Local Template¶
[Cfg] Each DUT runs L2BD switching with VLAN and 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
And Initialize L2 bridge domains with Vhost-User and VLAN in circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.21.2. 64B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.21.3. 64B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.21.4. 64B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.21.5. 1518B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.21.6. 1518B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.21.7. 1518B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.21.8. 9000B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.21.9. 9000B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.21.10. 9000B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.21.11. IMIX-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.21.12. IMIX-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.21.13. IMIX-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.22. 2n1l-10ge2p1x710-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD with vhost and IEEE 802.1Q test cases
[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. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[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 l2xc 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 | DOT1Q | L2BDMACLRN | BASE | VHOST | 1VM
| ... | VHOST_1024 | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.22.1. Local Template¶
[Cfg] Each DUT runs L2BD switching with VLAN and 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
And Initialize L2 bridge domains with Vhost-User and VLAN in circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.22.2. 64B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.22.3. 64B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.22.4. 64B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.22.5. 1518B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.22.6. 1518B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.22.7. 1518B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.22.8. 9000B-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.22.9. 9000B-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.22.10. 9000B-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.22.11. IMIX-1c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.22.12. IMIX-2c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.22.13. IMIX-4c-dot1q-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.23. 2n1l-10ge2p1x710-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD with vhost and IEEE 802.1Q test cases
[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. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[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 l2xc interconnecting virtio 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 | DOT1Q | L2BDMACLRN | BASE | VIRTIO | 1VM
| ... | VIRTIO_1024 | NF_VPPL2XC_VIRTIO | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${bd_id1}= | 1
| ${bd_id2}= | 2
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.23.1. Local Template¶
[Cfg] Each DUT runs L2BD switching with VLAN and 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
And Initialize L2 bridge domains with Vhost-User and VLAN in circular topology ${bd_id1} ${bd_id2} ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2virtiovr1024
Then Find NDR and PDR intervals using optimized search
7.13.23.2. 64B-1c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.23.3. 64B-2c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.23.4. 64B-4c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.23.5. 1518B-1c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.23.6. 1518B-2c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.23.7. 1518B-4c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.23.8. 9000B-1c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.23.9. 9000B-2c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.23.10. 9000B-4c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.23.11. IMIX-1c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.23.12. IMIX-2c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.23.13. IMIX-4c-dot1q-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.24. 2n1l-10ge2p1x710-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC with vhost abd IEEE 802.1Q test cases
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect of IPv4. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[Cfg] DUT configuration: DUT1 is configured with L2 cross-connect and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd 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 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.24.1. Local Template¶
[Cfg] Each DUT runs L2XC switching with VLAN and 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
And Initialize L2 xconnect with Vhost-User and VLAN in circular topology ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.24.2. 64B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.24.3. 64B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.24.4. 64B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.24.5. 1518B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.24.6. 1518B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.24.7. 1518B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.24.8. 9000B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.24.9. 9000B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.24.10. 9000B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.24.11. IMIX-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.24.12. IMIX-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.24.13. IMIX-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.25. 2n1l-10ge2p1x710-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC with vhost abd IEEE 802.1Q test cases
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect of IPv4. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[Cfg] DUT configuration: DUT1 is configured with L2 cross-connect and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc 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 | DOT1Q | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.25.1. Local Template¶
[Cfg] Each DUT runs L2XC switching with VLAN and 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
And Initialize L2 xconnect with Vhost-User and VLAN in circular topology ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.25.2. 64B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.25.3. 64B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.25.4. 64B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.25.5. 1518B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.25.6. 1518B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.25.7. 1518B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.25.8. 9000B-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.25.9. 9000B-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.25.10. 9000B-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.25.11. IMIX-1c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.25.12. IMIX-2c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.25.13. IMIX-4c-dot1q-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.26. 2n1l-10ge2p1x710-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC with vhost abd IEEE 802.1Q test cases
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for L2 cross connect of IPv4. IEEE 802.1Q tagging is applied on link between DUT1-if2 and TG-if2.
[Cfg] DUT configuration: DUT1 is configured with L2 cross-connect and MAC learning enabled. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting virtio 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 | DOT1Q | L2XCFWD | BASE | VIRTIO | 1VM | VIRTIO_1024
| ... | NF_VPPL2XC_VIRTIO | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc
|
| 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 | perfmon_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}= | ${4}
| ${subid}= | 10
| ${tag_rewrite}= | pop-1
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-dot1qip4asym-ip4src2547.13.26.1. Local Template¶
[Cfg] Each DUT runs L2XC switching with VLAN and 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
And Initialize L2 xconnect with Vhost-User and VLAN in circular topology ${subid} ${tag_rewrite}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=vppl2xc_2virtiovr1024
Then Find NDR and PDR intervals using optimized search
7.13.26.2. 64B-1c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.26.3. 64B-2c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.26.4. 64B-4c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.26.5. 1518B-1c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.26.6. 1518B-2c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.26.7. 1518B-4c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.26.8. 9000B-1c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.26.9. 9000B-2c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.26.10. 9000B-4c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.26.11. IMIX-1c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.26.12. IMIX-2c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.26.13. IMIX-4c-dot1q-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.27. 2n1l-10ge2p1x710-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost
[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 testpmd 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 the TG node 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 | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.27.1. 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=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.27.2. 64B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.27.3. 64B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.27.4. 64B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.27.5. 1518B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.27.6. 1518B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.27.7. 1518B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.27.8. 9000B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.27.9. 9000B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.27.10. 9000B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.27.11. IMIX-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.27.12. IMIX-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.27.13. IMIX-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.28. 2n1l-10ge2p1x710-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost
[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 l2xc 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 the TG node 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 | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.28.1. 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=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.28.2. 64B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.28.3. 64B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.28.4. 64B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.28.5. 1518B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.28.6. 1518B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.28.7. 1518B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.28.8. 9000B-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.28.9. 9000B-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.28.10. 9000B-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.28.11. IMIX-1c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.28.12. IMIX-2c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.28.13. IMIX-4c-eth-l2bdbasemaclrn-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.29. 2n1l-10ge2p1x710-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2BD test cases with vhost
[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 l2xc interconnecting virtio 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 the TG node 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 | VIRTIO | 1VM | VIRTIO_1024
| ... | NF_VPPL2XC_VIRTIO | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.29.1. 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=${True} vnf=vppl2xc_2virtiovr1024
Then Find NDR and PDR intervals using optimized search
7.13.29.2. 64B-1c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.29.3. 64B-2c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.29.4. 64B-4c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.29.5. 1518B-1c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.29.6. 1518B-2c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.29.7. 1518B-4c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.29.8. 9000B-1c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.29.9. 9000B-2c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.29.10. 9000B-4c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.29.11. IMIX-1c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.29.12. IMIX-2c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.29.13. IMIX-4c-eth-l2bdbasemaclrn-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.30. 2n1l-10ge2p1x710-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[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 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd 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 the TG node 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 | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2xcbase-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.30.1. Local Template¶
[Cfg] DUT runs L2XC 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 ${nic_driver}
And Initialize layer interface
And Initialize L2 xconnect with Vhost-User nf_nodes=${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=${True} vnf=testpmd_io
Then Find NDR and PDR intervals using optimized search
7.13.30.2. 64B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.30.3. 64B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.30.4. 64B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.30.5. 1518B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.30.6. 1518B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.30.7. 1518B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.30.8. 9000B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.30.9. 9000B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.30.10. 9000B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.30.11. IMIX-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.30.12. IMIX-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.30.13. IMIX-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.31. 2n1l-10ge2p1x710-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[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 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc 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 the TG node 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 | L2XCFWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPL2XC | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.31.1. Local Template¶
[Cfg] DUT runs L2XC 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 ${nic_driver}
And Initialize layer interface
And Initialize L2 xconnect with Vhost-User nf_nodes=${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=${True} vnf=vppl2xc_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.31.2. 64B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.31.3. 64B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.31.4. 64B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.31.5. 1518B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.31.6. 1518B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.31.7. 1518B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.31.8. 9000B-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.31.9. 9000B-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.31.10. 9000B-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.31.11. IMIX-1c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.31.12. IMIX-2c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.31.13. IMIX-4c-eth-l2xcbase-eth-2vhostvr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.32. 2n1l-10ge2p1x710-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr suite¶
RFC2544: Pkt throughput L2XC test cases with vhost
[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 cross- connect. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP l2xc interconnecting virtio 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 the TG node 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 | L2XCFWD | BASE | VIRTIO | 1VM | VIRTIO_1024
| ... | NF_VPPL2XC_VIRTIO | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc
|
| 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 | perfmon_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-2n-ethip4-ip4src2547.13.32.1. Local Template¶
[Cfg] DUT runs L2XC 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 ${nic_driver}
And Initialize layer interface
And Initialize L2 xconnect with Vhost-User nf_nodes=${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=${True} vnf=vppl2xc_2virtiovr1024
Then Find NDR and PDR intervals using optimized search
7.13.32.2. 64B-1c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.32.3. 64B-2c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.32.4. 64B-4c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.32.5. 1518B-1c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.32.6. 1518B-2c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.32.7. 1518B-4c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.32.8. 9000B-1c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.32.9. 9000B-2c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.32.10. 9000B-4c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.32.11. IMIX-1c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.32.12. IMIX-2c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.32.13. IMIX-4c-eth-l2xcbase-eth-2virtiovr1024-1vm-vppl2xc-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.33. 2n1l-10ge2p1x710-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr suite¶
RFC2544: Pkt throughput IPv4 test cases with vhost
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for IPv4 routing.
[Cfg] DUT configuration: DUT1 is configured with IPv4. routing and two static IPv4 /24 route entries. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running testpmd 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, 253 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 the TG node 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 | IP4FWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_TESTPMD | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4-ip4base-eth-2vhostvr1024-1vm
|
| 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 | perfmon_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}= | L3
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-ethip4-ip4src2537.13.33.1. Local Template¶
[Cfg] DUT runs IPv4 routing 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 ${nic_driver}
And Initialize layer interface
And Initialize IPv4 forwarding with vhost in 2-node circular topology nf_nodes=${nf_nodes} testpmd_mac=${TRUE}
And Configure chains of NFs connected via vhost-user nf_chains=${nf_chains} nf_nodes=${nf_nodes} jumbo=${jumbo} use_tuned_cfs=${False} auto_scale=${True} vnf=testpmd_mac
Then Find NDR and PDR intervals using optimized search
7.13.33.2. 64B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.33.3. 64B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.33.4. 64B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.33.5. 1518B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.33.6. 1518B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.33.7. 1518B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.33.8. 9000B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.33.9. 9000B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.33.10. 9000B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.33.11. IMIX-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.33.12. IMIX-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.33.13. IMIX-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.34. 2n1l-10ge2p1x710-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr suite¶
RFC2544: Pkt throughput IPv4 test cases with vhost
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for IPv4 routing.
[Cfg] DUT configuration: DUT1 is configured with IPv4. routing and two static IPv4 /24 route entries. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP ip4 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, 253 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 the TG node 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 | IP4FWD | BASE | VHOST | 1VM | VHOST_1024
| ... | NF_VPPIP4 | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4-ip4base-eth-2vhostvr1024-1vm-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 | perfmon_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}= | L3
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-ethip4-ip4src2537.13.34.1. Local Template¶
[Cfg] DUT runs IPv4 routing 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 ${nic_driver}
And Initialize layer interface
And Initialize IPv4 forwarding with vhost in 2-node circular topology nf_nodes=${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=${True} vnf=vppip4noarp_2vhostvr1024
Then Find NDR and PDR intervals using optimized search
7.13.34.2. 64B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.34.3. 64B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.34.4. 64B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.34.5. 1518B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.34.6. 1518B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.34.7. 1518B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.34.8. 9000B-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.34.9. 9000B-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.34.10. 9000B-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.34.11. IMIX-1c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.34.12. IMIX-2c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.34.13. IMIX-4c-ethip4-ip4base-eth-2vhostvr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}
7.13.35. 2n1l-10ge2p1x710-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr suite¶
RFC2544: Pkt throughput IPv4 test cases with vhost
[Top] Network Topologies: TG-DUT1-TG 2-node circular topology with single links between nodes.
[Enc] Packet Encapsulations: Eth-IPv4 for IPv4 routing.
[Cfg] DUT configuration: DUT1 is configured with IPv4. routing and two static IPv4 /24 route entries. Qemu VNFs are connected to VPP via vhost-user interfaces. Guest is running VPP ip4 interconnecting virtio 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, 253 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 the TG node 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 | IP4FWD | BASE | VIRTIO | 1VM | VIRTIO_1024
| ... | NF_VPPIP4_VIRTIO | DRV_VFIO_PCI
| ... | RXQ_SIZE_0 | TXQ_SIZE_0
| ... | ethip4-ip4base-eth-2virtiovr1024-1vm-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 | perfmon_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}= | L3
| ${overhead}= | ${0}
| ${nf_dtcr}= | ${1}
| ${nf_dtc}= | ${1}
| ${nf_chains}= | ${1}
| ${nf_nodes}= | ${1}
# Traffic profile:
| ${traffic_profile}= | trex-stl-2n-ethip4-ip4src2537.13.35.1. Local Template¶
[Cfg] DUT runs IPv4 routing 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 ${nic_driver}
And Initialize layer interface
And Initialize IPv4 forwarding with vhost in 2-node circular topology nf_nodes=${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=${True} vnf=vppip4noarp_2virtiovr1024
Then Find NDR and PDR intervals using optimized search
7.13.35.2. 64B-1c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${1}
7.13.35.3. 64B-2c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${2}
7.13.35.4. 64B-4c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${64} phy_cores=${4}
7.13.35.5. 1518B-1c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${1}
7.13.35.6. 1518B-2c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${2}
7.13.35.7. 1518B-4c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${1518} phy_cores=${4}
7.13.35.8. 9000B-1c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${1}
7.13.35.9. 9000B-2c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${2}
7.13.35.10. 9000B-4c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=${9000} phy_cores=${4}
7.13.35.11. IMIX-1c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${1}
7.13.35.12. IMIX-2c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${2}
7.13.35.13. IMIX-4c-ethip4-ip4base-eth-2virtiovr1024-1vm-vppip4-ndrpdr¶
frame_size=IMIX_v4_1 phy_cores=${4}