Hi I run a stress test on stateful VPP. After a while it crash and raise an os panic signal. I run this test on both single and multicore vpp and every time it happens.
My Scenario: 1. VPP: Default startup.conf configuration (main and worker on same core) Session timeouts are configured as follows: vppctl set acl-plugin session timeout udp idle 5 vppctl set acl-plugin session timeout tcp idle 10 vppctl set acl-plugin session timeout tcp transient 5 2 interface in brdige mode add permit + reflect acl rule on both Input & Output of all interfaces Trex: ./t-rex-64 -f cap2/my_custom_test.yaml --cfg cfg/concurrent_connection_trex_config.yaml -c 8 -m 3000 2. VPP: 1 main core + 1 worker core Session timeouts are configured as follows: vppctl set acl-plugin session timeout udp idle 5 vppctl set acl-plugin session timeout tcp idle 10 vppctl set acl-plugin session timeout tcp transient 5 2 interface in brdige mode add permit + reflect acl rule on both Input & Output of all interfaces Trex: ./t-rex-64 -f cap2/my_custom_test.yaml --cfg cfg/concurrent_connection_trex_config.yaml -c 8 -m 3000 3. VPP: 1 main core + 4 worker core Session timeouts are configured as follows: vppctl set acl-plugin session timeout udp idle 5 vppctl set acl-plugin session timeout tcp idle 10 vppctl set acl-plugin session timeout tcp transient 5 2 interface in brdige mode add permit + reflect acl rule on both Input & Output of all interfaces Trex: ./t-rex-64 -f cap2/my_custom_test.yaml --cfg cfg/concurrent_connection_trex_config.yaml -c 8 -m 3000 I also attached my Trex yaml and config files. Sent from Outlook<http://aka.ms/weboutlook>
my_custom_test.yaml
Description: my_custom_test.yaml
new_connection_test_pcap.pcap
Description: new_connection_test_pcap.pcap
concurrent_connection_trex_config.yaml
Description: concurrent_connection_trex_config.yaml
_______________________________________________ vpp-dev mailing list vpp-dev@lists.fd.io https://lists.fd.io/mailman/listinfo/vpp-dev
