diff options
| -rw-r--r-- | resources/libraries/python/TrafficGenerator.py | 45 | ||||
| -rw-r--r-- | resources/libraries/python/VppConfigGenerator.py | 6 | ||||
| -rw-r--r-- | resources/libraries/robot/performance.robot | 40 | ||||
| -rwxr-xr-x | resources/tools/t-rex/t-rex-stateless.py | 74 | ||||
| -rw-r--r-- | tests/perf/Long_IPv6_Fib_200k_Intel-X520-DA2.robot | 492 | ||||
| -rw-r--r-- | tests/perf/Long_IPv6_Fib_20k_Intel-X520-DA2.robot | 492 | ||||
| -rw-r--r-- | tests/perf/Long_IPv6_Fib_2M_Intel-X520-DA2.robot | 492 | ||||
| -rw-r--r-- | tests/perf/Short_IPv6_Fib_200k_Intel-X520-DA2.robot | 213 | ||||
| -rw-r--r-- | tests/perf/Short_IPv6_Fib_20k_Intel-X520-DA2.robot | 213 | ||||
| -rw-r--r-- | tests/perf/Short_IPv6_Fib_2M_Intel-X520-DA2.robot | 213 |
10 files changed, 2256 insertions, 24 deletions
diff --git a/resources/libraries/python/TrafficGenerator.py b/resources/libraries/python/TrafficGenerator.py index a640c1d80a..28a75669b2 100644 --- a/resources/libraries/python/TrafficGenerator.py +++ b/resources/libraries/python/TrafficGenerator.py @@ -406,6 +406,51 @@ class TrafficGenerator(object): _p0, _p1, _async, _latency, warmup_time), timeout=int(duration)+60) + elif traffic_type in ["3-node-IPv6-dst-10000"]: + (ret, stdout, stderr) = ssh.exec_command( + "sh -c '{0}/resources/tools/t-rex/t-rex-stateless.py " + "--duration={1} -r {2} -s {3} -6 " + "--p{4}_src_start_ip 2001:1::1 " + "--p{4}_dst_start_ip 2001:2::0 " + "--p{4}_dst_end_ip 2001:2::270F " + "--p{5}_src_start_ip 2001:2::1 " + "--p{5}_dst_start_ip 2001:1::0 " + "--p{5}_dst_end_ip 2001:1::270F " + "{6} {7} --warmup_time={8}'".format(Constants.REMOTE_FW_DIR, + duration, rate, framesize, + _p0, _p1, _async, _latency, + warmup_time), + timeout=int(duration)+60) + elif traffic_type in ["3-node-IPv6-dst-100000"]: + (ret, stdout, stderr) = ssh.exec_command( + "sh -c '{0}/resources/tools/t-rex/t-rex-stateless.py " + "--duration={1} -r {2} -s {3} -6 " + "--p{4}_src_start_ip 2001:1::1 " + "--p{4}_dst_start_ip 2001:2::0 " + "--p{4}_dst_end_ip 2001:2::1:869F " + "--p{5}_src_start_ip 2001:2::1 " + "--p{5}_dst_start_ip 2001:1::0 " + "--p{5}_dst_end_ip 2001:1::1:869F " + "{6} {7} --warmup_time={8}'".format(Constants.REMOTE_FW_DIR, + duration, rate, framesize, + _p0, _p1, _async, _latency, + warmup_time), + timeout=int(duration)+60) + elif traffic_type in ["3-node-IPv6-dst-1000000"]: + (ret, stdout, stderr) = ssh.exec_command( + "sh -c '{0}/resources/tools/t-rex/t-rex-stateless.py " + "--duration={1} -r {2} -s {3} -6 " + "--p{4}_src_start_ip 2001:1::1 " + "--p{4}_dst_start_ip 2001:2::0 " + "--p{4}_dst_end_ip 2001:2::F:423F " + "--p{5}_src_start_ip 2001:2::1 " + "--p{5}_dst_start_ip 2001:1::0 " + "--p{5}_dst_end_ip 2001:1::F:423F " + "{6} {7} --warmup_time={8}'".format(Constants.REMOTE_FW_DIR, + duration, rate, framesize, + _p0, _p1, _async, _latency, + warmup_time), + timeout=int(duration)+60) else: raise NotImplementedError('Unsupported traffic type') diff --git a/resources/libraries/python/VppConfigGenerator.py b/resources/libraries/python/VppConfigGenerator.py index de1f0c7e5c..ad3f23483e 100644 --- a/resources/libraries/python/VppConfigGenerator.py +++ b/resources/libraries/python/VppConfigGenerator.py @@ -60,6 +60,12 @@ dpdk {{ {nomultiseg} {enablevhostuser} }} + +ip6 {{ + hash-buckets 2000000 + heap-size 500m +}} + """ # End VPP configuration template. diff --git a/resources/libraries/robot/performance.robot b/resources/libraries/robot/performance.robot index 78cdd8b845..bb9e8ab135 100644 --- a/resources/libraries/robot/performance.robot +++ b/resources/libraries/robot/performance.robot @@ -386,6 +386,46 @@ | | Vpp Route Add | ${dut1} | 2001:2::0 | ${prefix} | 2001:3::2 | ${dut1_if2} | | Vpp Route Add | ${dut2} | 2001:1::0 | ${prefix} | 2001:3::1 | ${dut2_if1} +| Scale IPv6 forwarding initialized in a 3-node circular topology +| | [Documentation] +| | ... | Custom setup of IPv6 topology with scalability of ip routes on all +| | ... | DUT nodes in 3-node circular topology +| | ... +| | ... | *Arguments:* +| | ... | - ${count} - IP route count. Type: integer +| | ... +| | ... | *Return:* +| | ... | - No value returned +| | ... +| | ... | *Example:* +| | ... +| | ... | \| Scale IPv6 forwarding initialized in a 3-node circular topology \ +| | ... | \| 100000 \| +| | [Arguments] | ${count} +| | ${subn_prefix}= | Set Variable | 64 +| | ${host_prefix}= | Set Variable | 128 +| | VPP Set If IPv6 Addr | ${dut1} | ${dut1_if1} | 2001:3::1 | ${subn_prefix} +| | VPP Set If IPv6 Addr | ${dut1} | ${dut1_if2} | 2001:4::1 | ${subn_prefix} +| | VPP Set If IPv6 Addr | ${dut2} | ${dut2_if1} | 2001:4::2 | ${subn_prefix} +| | VPP Set If IPv6 Addr | ${dut2} | ${dut2_if2} | 2001:5::1 | ${subn_prefix} +| | ${tg1_if1_mac}= | Get Interface MAC | ${tg} | ${tg_if1} +| | ${tg1_if2_mac}= | Get Interface MAC | ${tg} | ${tg_if2} +| | ${dut1_if2_mac}= | Get Interface MAC | ${dut1} | ${dut1_if2} +| | ${dut2_if1_mac}= | Get Interface MAC | ${dut2} | ${dut2_if1} +| | Vpp nodes ra suppress link layer | ${nodes} +| | Add Ip Neighbor | ${dut1} | ${dut1_if1} | 2001:3::2 | ${tg1_if1_mac} +| | Add Ip Neighbor | ${dut1} | ${dut1_if2} | 2001:4::2 | ${dut2_if1_mac} +| | Add Ip Neighbor | ${dut2} | ${dut2_if1} | 2001:4::1 | ${dut1_if2_mac} +| | Add Ip Neighbor | ${dut2} | ${dut2_if2} | 2001:5::2 | ${tg1_if2_mac} +| | Vpp Route Add | ${dut1} | 2001:2::0 | ${host_prefix} | 2001:4::2 +| | ... | interface=${dut1_if2} | count=${count} +| | Vpp Route Add | ${dut1} | 2001:1::0 | ${host_prefix} | 2001:3::2 +| | ... | interface=${dut1_if1} | count=${count} +| | Vpp Route Add | ${dut2} | 2001:1::0 | ${host_prefix} | 2001:4::1 +| | ... | interface=${dut2_if2} | count=${count} +| | Vpp Route Add | ${dut2} | 2001:2::0 | ${host_prefix} | 2001:5::2 +| | ... | interface=${dut2_if1} | count=${count} + | L2 xconnect initialized in a 3-node circular topology | | [Documentation] | | ... | Setup L2 xconnect topology by cross connecting two interfaces on diff --git a/resources/tools/t-rex/t-rex-stateless.py b/resources/tools/t-rex/t-rex-stateless.py index 9ee34f6c22..6c7579d0c3 100755 --- a/resources/tools/t-rex/t-rex-stateless.py +++ b/resources/tools/t-rex/t-rex-stateless.py @@ -96,7 +96,7 @@ def get_start_end_ipv6(start_ip, end_ip): print "IPv6: start_ip is greater then end_ip" sys.exit(2) - max_p1 = abs(int(lo1) - int(lo2)) + 1 + max_p1 = abs(int(lo1) - int(lo2)) base_p1 = lo1 except AddressValueError as ex_error: print ex_error @@ -259,33 +259,59 @@ def create_streams_v6(traffic_options, frame_size=78): p2_src_end_ip = traffic_options['p2_src_end_ip'] p2_dst_start_ip = traffic_options['p2_dst_start_ip'] - base_p1, max_p1 = get_start_end_ipv6(p1_src_start_ip, p1_src_end_ip) - base_p2, max_p2 = get_start_end_ipv6(p2_src_start_ip, p2_src_end_ip) + p1_dst_end_ip = traffic_options['p1_dst_end_ip'] + p2_dst_end_ip = traffic_options['p2_dst_end_ip'] base_pkt_a = Ether()/IPv6(src=p1_src_start_ip, dst=p1_dst_start_ip) base_pkt_b = Ether()/IPv6(src=p2_src_start_ip, dst=p2_dst_start_ip) - # The following code applies raw instructions to packet (IP src increment). - # It splits the generated traffic by "ip_src" variable to cores - vm1 = STLScVmRaw([STLVmFlowVar(name="ipv6_src", - min_value=base_p1, - max_value=max_p1+base_p1, - size=8, op="inc"), - STLVmWrFlowVar(fv_name="ipv6_src", pkt_offset="IPv6.src", - offset_fixup=8) - ] - , split_by_field="ipv6_src") - - # The following code applies raw instructions to packet (IP src increment). - # It splits the generated traffic by "ip_src" variable to cores - vm2 = STLScVmRaw([STLVmFlowVar(name="ipv6_src", - min_value=base_p2, - max_value=max_p2+base_p2, - size=8, op="inc"), - STLVmWrFlowVar(fv_name="ipv6_src", pkt_offset="IPv6.src", - offset_fixup=8) - ] - , split_by_field="ipv6_src") + # The following code applies raw instructions to packet (IP src/dst + # increment). It splits the generated traffic by "ip_src"/"ip_dst" variable + # to cores. + if p1_dst_end_ip and p2_dst_end_ip: + base_p1, max_p1 = get_start_end_ipv6(p1_dst_start_ip, p1_dst_end_ip) + base_p2, max_p2 = get_start_end_ipv6(p2_dst_start_ip, p2_dst_end_ip) + + vm1 = STLScVmRaw([STLVmFlowVar(name="ipv6_dst", + min_value=base_p1, + max_value=max_p1+base_p1, + size=8, op="inc"), + STLVmWrFlowVar(fv_name="ipv6_dst", + pkt_offset="IPv6.dst", + offset_fixup=8) + ] + , split_by_field="ipv6_dst") + vm2 = STLScVmRaw([STLVmFlowVar(name="ipv6_dst", + min_value=base_p2, + max_value=max_p2+base_p2, + size=8, op="inc"), + STLVmWrFlowVar(fv_name="ipv6_dst", + pkt_offset="IPv6.dst", + offset_fixup=8) + ] + , split_by_field="ipv6_dst") + else: + base_p1, max_p1 = get_start_end_ipv6(p1_src_start_ip, p1_src_end_ip) + base_p2, max_p2 = get_start_end_ipv6(p2_src_start_ip, p2_src_end_ip) + + vm1 = STLScVmRaw([STLVmFlowVar(name="ipv6_src", + min_value=base_p1, + max_value=max_p1+base_p1, + size=8, op="inc"), + STLVmWrFlowVar(fv_name="ipv6_src", + pkt_offset="IPv6.src", + offset_fixup=8) + ] + , split_by_field="ipv6_src") + vm2 = STLScVmRaw([STLVmFlowVar(name="ipv6_src", + min_value=base_p2, + max_value=max_p2+base_p2, + size=8, op="inc"), + STLVmWrFlowVar(fv_name="ipv6_src", + pkt_offset="IPv6.src", + offset_fixup=8) + ] + , split_by_field="ipv6_src") return create_streams_v46(base_pkt_a, base_pkt_b, vm1, vm2, frame_size) diff --git a/tests/perf/Long_IPv6_Fib_200k_Intel-X520-DA2.robot b/tests/perf/Long_IPv6_Fib_200k_Intel-X520-DA2.robot new file mode 100644 index 0000000000..470179e33a --- /dev/null +++ b/tests/perf/Long_IPv6_Fib_200k_Intel-X520-DA2.robot @@ -0,0 +1,492 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_LONG +| ... | FIB_200K | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *RFC2544: Pkt throughput IPv6 routing test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x100k static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* TG finds and reports throughput NDR (Non Drop +| ... | Rate) with zero packet loss tolerance or 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 either binary search or linear search +| ... | algorithms with configured starting rate and final step that determines +| ... | throughput measurement resolution. Test packets are generated by TG on +| ... | links to DUTs. TG traffic profile contains two L3 flow-groups +| ... | (flow-group per direction, 100k flows per flow-group) with all packets +| ... | containing Ethernet header, IPv6 header withq IP and static payload. +| ... | MAC addresses are matching MAC addresses of the TG node interfaces. +| ... | Incrementing of IP.dst (IPv6 destination address) field is applied to +| ... | both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 100000 + +*** Test Cases *** +| TC01: 78B NDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC02: 78B PDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC03: 1518B NDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC04: 1518B PDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC05: 9000B NDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC06: 9000B PDR binary search - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC07: 78B NDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC08: 78B PDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC09: 1518B NDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC10: 1518B PDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC11: 9000B NDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC12: 9000B PDR binary search - DUT IPv6 Fib 2x100k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC13: 78B NDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC14: 78B PDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC15: 1518B NDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC16: 1518B PDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC17: 9000B NDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC18: 9000B PDR binary search - DUT IPv6 Fib 2x100k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} diff --git a/tests/perf/Long_IPv6_Fib_20k_Intel-X520-DA2.robot b/tests/perf/Long_IPv6_Fib_20k_Intel-X520-DA2.robot new file mode 100644 index 0000000000..0345939f76 --- /dev/null +++ b/tests/perf/Long_IPv6_Fib_20k_Intel-X520-DA2.robot @@ -0,0 +1,492 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_LONG +| ... | FIB_20K | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *RFC2544: Pkt throughput IPv6 routing test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x10k static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* TG finds and reports throughput NDR (Non Drop +| ... | Rate) with zero packet loss tolerance or 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 either binary search or linear search +| ... | algorithms with configured starting rate and final step that determines +| ... | throughput measurement resolution. Test packets are generated by TG on +| ... | links to DUTs. TG traffic profile contains two L3 flow-groups +| ... | (flow-group per direction, 10k flows per flow-group) with all packets +| ... | containing Ethernet header, IPv6 header withq IP and static payload. +| ... | MAC addresses are matching MAC addresses of the TG node interfaces. +| ... | Incrementing of IP.dst (IPv6 destination address) field is applied to +| ... | both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 10000 + +*** Test Cases *** +| TC01: 78B NDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC02: 78B PDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC03: 1518B NDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC04: 1518B PDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC05: 9000B NDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC06: 9000B PDR binary search - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC07: 78B NDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC08: 78B PDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC09: 1518B NDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC10: 1518B PDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC11: 9000B NDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC12: 9000B PDR binary search - DUT IPv6 Fib 2x10k - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC13: 78B NDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC14: 78B PDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC15: 1518B NDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC16: 1518B PDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC17: 9000B NDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC18: 9000B PDR binary search - DUT IPv6 Fib 2x10k - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} diff --git a/tests/perf/Long_IPv6_Fib_2M_Intel-X520-DA2.robot b/tests/perf/Long_IPv6_Fib_2M_Intel-X520-DA2.robot new file mode 100644 index 0000000000..fa4bab241f --- /dev/null +++ b/tests/perf/Long_IPv6_Fib_2M_Intel-X520-DA2.robot @@ -0,0 +1,492 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_LONG +| ... | FIB_2M | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *RFC2544: Pkt throughput IPv6 routing test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x1M static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* TG finds and reports throughput NDR (Non Drop +| ... | Rate) with zero packet loss tolerance or 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 either binary search or linear search +| ... | algorithms with configured starting rate and final step that determines +| ... | throughput measurement resolution. Test packets are generated by TG on +| ... | links to DUTs. TG traffic profile contains two L3 flow-groups +| ... | (flow-group per direction, 1M flows per flow-group) with all packets +| ... | containing Ethernet header, IPv6 header withq IP and static payload. +| ... | MAC addresses are matching MAC addresses of the TG node interfaces. +| ... | Incrementing of IP.dst (IPv6 destination address) field is applied to +| ... | both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 1000000 + +*** Test Cases *** +| TC01: 78B NDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC02: 78B PDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC03: 1518B NDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC04: 1518B PDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC05: 9000B NDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC06: 9000B PDR binary search - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | PDR +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC07: 78B NDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC08: 78B PDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC09: 1518B NDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC10: 1518B PDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC11: 9000B NDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC12: 9000B PDR binary search - DUT IPv6 Fib 2x1M - 2threads 2cores 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 threads, 2 phy cores, \ +| | ... | 1 receive queue per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC13: 78B NDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC14: 78B PDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 78 Byte frames +| | ... | using binary search start at 10GE linerate, step 100kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR +| | ${framesize}= | Set Variable | 78 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_78B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC15: 1518B NDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 100000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC16: 1518B PDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 1518 Byte frames +| | ... | using binary search start at 10GE linerate, step 10kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 1518 +| | ${min_rate}= | Set Variable | 10000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_1518B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} + +| TC17: 9000B NDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find NDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find NDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} + +| TC18: 9000B PDR binary search - DUT IPv6 Fib 2x1M - 4threads 4cores 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 threads, 4 phy cores, \ +| | ... | 2 receive queues per NIC port. [Ver] Find PDR for 9000 Byte frames +| | ... | using binary search start at 10GE linerate, step 5kpps, LT=0.5%. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | PDR | SKIP_PATCH +| | ${framesize}= | Set Variable | 9000 +| | ${min_rate}= | Set Variable | 5000 +| | ${max_rate}= | Set Variable | ${10Ge_linerate_pps_9000B} +| | ${binary_min}= | Set Variable | ${min_rate} +| | ${binary_max}= | Set Variable | ${max_rate} +| | ${threshold}= | Set Variable | ${min_rate} +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Find PDR using binary search and pps | ${framesize} | ${binary_min} +| | ... | ${binary_max} +| | ... | 3-node-IPv6-dst-${rts_per_flow} +| | ... | ${min_rate} | ${max_rate} +| | ... | ${threshold} +| | ... | ${glob_loss_acceptance} +| | ... | ${glob_loss_acceptance_type} diff --git a/tests/perf/Short_IPv6_Fib_200k_Intel-X520-DA2.robot b/tests/perf/Short_IPv6_Fib_200k_Intel-X520-DA2.robot new file mode 100644 index 0000000000..ec09bb8d02 --- /dev/null +++ b/tests/perf/Short_IPv6_Fib_200k_Intel-X520-DA2.robot @@ -0,0 +1,213 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_SHORT +| ... | FIB_200K | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *Reference NDR throughput IPv6 routing verify test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x100k static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* In short performance tests, TG verifies +| ... | DUTs' throughput at ref-NDR (reference Non Drop Rate) with zero packet +| ... | loss tolerance. Ref-NDR value is periodically updated acording to +| ... | formula: ref-NDR = 0.9x NDR, where NDR is found in RFC2544 long +| ... | performance tests for the same DUT confiiguration. Test packets are +| ... | generated by TG on links to DUTs. TG traffic profile contains two L3 +| ... | flow-groups (flow-group per direction, 100k flows per flow-group) with +| ... | all packets containing Ethernet header, IPv6 header and generated +| ... | payload. Incrementing of IP.dst (IPv6 destination address) field is +| ... | applied to both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 100000 + +*** Test Cases *** +| TC01: Verify 78B ref-NDR at 2x 2.9Mpps - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 2.9mpps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC02: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC03: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x100k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC04: Verify 78B ref-NDR at 2x 5.9Mpps - DUT IPv6 Fib 2x100k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 5.9mpps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC05: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x100k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC06: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x100k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC07: Verify 78B ref-NDR at 2x 10.3Mpps - DUT IPv6 Fib 2x100k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 10.3mpps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC08: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x100k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC09: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x100k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} diff --git a/tests/perf/Short_IPv6_Fib_20k_Intel-X520-DA2.robot b/tests/perf/Short_IPv6_Fib_20k_Intel-X520-DA2.robot new file mode 100644 index 0000000000..1981af0d8e --- /dev/null +++ b/tests/perf/Short_IPv6_Fib_20k_Intel-X520-DA2.robot @@ -0,0 +1,213 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_SHORT +| ... | FIB_20K | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *Reference NDR throughput IPv6 routing verify test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x10k static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* In short performance tests, TG verifies +| ... | DUTs' throughput at ref-NDR (reference Non Drop Rate) with zero packet +| ... | loss tolerance. Ref-NDR value is periodically updated acording to +| ... | formula: ref-NDR = 0.9x NDR, where NDR is found in RFC2544 long +| ... | performance tests for the same DUT confiiguration. Test packets are +| ... | generated by TG on links to DUTs. TG traffic profile contains two L3 +| ... | flow-groups (flow-group per direction, 10k flows per flow-group) with +| ... | all packets containing Ethernet header, IPv6 header and generated +| ... | payload. Incrementing of IP.dst (IPv6 destination address) field is +| ... | applied to both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 10000 + +*** Test Cases *** +| TC01: Verify 78B ref-NDR at 2x 2.9Mpps - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 2.9mpps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC02: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC03: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x10k - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC04: Verify 78B ref-NDR at 2x 5.9Mpps - DUT IPv6 Fib 2x10k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 5.9mpps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC05: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x10k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC06: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x10k - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC07: Verify 78B ref-NDR at 2x 10.0Mpps - DUT IPv6 Fib 2x10k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 10.0mpps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC08: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x10k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC09: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x10k - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} diff --git a/tests/perf/Short_IPv6_Fib_2M_Intel-X520-DA2.robot b/tests/perf/Short_IPv6_Fib_2M_Intel-X520-DA2.robot new file mode 100644 index 0000000000..782adc6598 --- /dev/null +++ b/tests/perf/Short_IPv6_Fib_2M_Intel-X520-DA2.robot @@ -0,0 +1,213 @@ +# Copyright (c) 2016 Cisco and/or its affiliates. +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at: +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +*** Settings *** +| Resource | resources/libraries/robot/performance.robot +| Force Tags | 3_NODE_SINGLE_LINK_TOPO | PERFTEST | HW_ENV | PERFTEST_SHORT +| ... | FIB_2M | NIC_Intel-X520-DA2 +| Suite Setup | 3-node Performance Suite Setup with DUT's NIC model +| ... | L3 | Intel-X520-DA2 +| Suite Teardown | 3-node Performance Suite Teardown +| Test Setup | Setup all DUTs before test +| Test Teardown | Run Keyword | Remove startup configuration of VPP from all DUTs +| Documentation | *Reference NDR throughput IPv6 routing verify test cases* +| ... +| ... | *[Top] Network Topologies:* TG-DUT1-DUT2-TG 3-node circular topology +| ... | with single links between nodes. +| ... | *[Enc] Packet Encapsulations:* Eth-IPv6 for IPv6 routing. +| ... | *[Cfg] DUT configuration:* DUT1 and DUT2 are configured with IPv6 +| ... | routing and 2x1M static IPv6 /64 route entries. DUT1 and DUT2 tested +| ... | with 2p10GE NIC X520 Niantic by Intel. +| ... | *[Ver] TG verification:* In short performance tests, TG verifies +| ... | DUTs' throughput at ref-NDR (reference Non Drop Rate) with zero packet +| ... | loss tolerance. Ref-NDR value is periodically updated acording to +| ... | formula: ref-NDR = 0.9x NDR, where NDR is found in RFC2544 long +| ... | performance tests for the same DUT confiiguration. Test packets are +| ... | generated by TG on links to DUTs. TG traffic profile contains two L3 +| ... | flow-groups (flow-group per direction, 1M flows per flow-group) with +| ... | all packets containing Ethernet header, IPv6 header and generated +| ... | payload. Incrementing of IP.dst (IPv6 destination address) field is +| ... | applied to both streams. +| ... | *[Ref] Applicable standard specifications:* RFC2544. + +*** Variables *** +| ${rts_per_flow}= | 1000000 + +*** Test Cases *** +| TC01: Verify 78B ref-NDR at 2x 2.3Mpps - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 2.3mpps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC02: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC03: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x1M - 1thread 1core 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 1 thread, 1 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 1_THREAD_NOHTT_RXQUEUES_1 | SINGLE_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '1' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC04: Verify 78B ref-NDR at 2x 4.9Mpps - DUT IPv6 Fib 2x1M - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 4.9mpps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC05: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x1M - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC06: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x1M - 2threads, 2cores, 1rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 2 thread, 2 phy core, \ +| | ... | 1 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 2_THREAD_NOHTT_RXQUEUES_1 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '2' worker threads and rxqueues '1' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC07: Verify 78B ref-NDR at 2x 9.5Mpps - DUT IPv6 Fib 2x1M - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 78 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 78 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 9.5mpps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC08: Verify 1518B ref-NDR at 2x 812.74kpps - DUT IPv6 Fib 2x1M - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 1518 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 1518 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 812743pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Add No Multi Seg to all DUTs +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} + +| TC09: Verify 9000B ref-NDR at 2x 138.58kpps - DUT IPv6 Fib 2x1M - 4threads, 4cores, 2rxq +| | [Documentation] +| | ... | [Cfg] DUT runs IPv6 routing config with 4 thread, 4 phy core, \ +| | ... | 2 receive queue per NIC port. [Ver] Verify ref-NDR for 9000 Byte +| | ... | frames using single trial throughput test. +| | [Tags] | 4_THREAD_NOHTT_RXQUEUES_2 | MULTI_THREAD | NDR +| | ${framesize}= | Set Variable | 9000 +| | ${duration}= | Set Variable | 10 +| | ${rate}= | Set Variable | 138580pps +| | Given Add '4' worker threads and rxqueues '2' without HTT to all DUTs +| | And Add PCI devices to DUTs from 3-node single link topology +| | And Apply startup configuration on all VPP DUTs +| | And Scale IPv6 forwarding initialized in a 3-node circular topology +| | ... | ${rts_per_flow} +| | Then Traffic should pass with no loss | ${duration} | ${rate} +| | ... | ${framesize} +| | ... | 3-node-IPv6-dst-${rts_per_flow} |