Since merging https://review.opendev.org/#/c/667547/, the relative frequency of test cases failures has changed. The frequency of test_connectivity_through_2_routers has decreased significantly. As a consequence. I did an analysis of frequency failures per test over the past 7 days (I went through the 500 occurrences that Kibana returns as a maximum. There may be more not visible beyond that 500 limit). This is what I found:
Since merging https:/ /review. opendev. org/#/c/ 667547/, the relative frequency of test cases failures has changed. The frequency of test_connectivi ty_through_ 2_routers has decreased significantly. As a consequence. I did an analysis of frequency failures per test over the past 7 days (I went through the 500 occurrences that Kibana returns as a maximum. There may be more not visible beyond that 500 limit). This is what I found:
test_qos_ basic_and_ update 48 dvr_to_ dvr_ha 39 dvr_to_ legacy 23 ty_through_ 2_routers 17 external_ ip 17 reachable_ through_ compute 10 subport_ lifecycle 9
test_from_
test_from_dvr_to_ha 38
test_from_
test_connectivi
test_snat_
test_vm_
test_trunk_
test_qos 8
So the big offenders as of now are test_qos_ basic_and_ update and the routers migrations