[ST] persistent kafka with 1 rep. instead ephemeral #9782
Conversation
Signed-off-by: hzrncik <[email protected]>
|
@strimzi-ci run tests --profile=all --env=STRIMZI_FEATURE_GATES=-KafkaNodePools,-UnidirectionalTopicOperator,-UseKRaft |
|
|
|
/azp run feature-gates-regression |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
Adding 2 more ZK nodes into STs could cause that our memory limit for current VMs, which we have can be reached. So I would be very careful (we can run regression pipeline more than 1 time to prove that it's okay). I think having some kind of probe to monitor memory/cpu usage would be good for this inside our STs |
|
/azp run regression |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
@see-quick yeah good suggestion, I have tried to avoid this solution in test case where more than 1 kafka is created, but still i wanted to create this situation at least in 2,3 tests to see if it it would really cause some resources problem as i did not hit this problem locally or in past. |
systemtest/src/test/java/io/strimzi/systemtest/mirrormaker/MirrorMaker2ST.java
Show resolved
Hide resolved
systemtest/src/test/java/io/strimzi/systemtest/operators/user/UserST.java
Show resolved
Hide resolved
…kafka clusters used Signed-off-by: hzrncik <[email protected]>
|
@strimzi-ci run tests --cluster-type=ocp --cluster-version=4.14 --install-type=bundle --profile=all --testcase=HttpBridgeST,MirrorMaker2ST,RackAwarenessST --parallel-count=1 --env=STRIMZI_FEATURE_GATES=+UnidirectionalTopicOperator;STRIMZI_USE_KRAFT_IN_TESTS=true |
|
|
|
/azp run kraft-regression |
|
Azure Pipelines successfully started running 1 pipeline(s). |
✔️ Test Summary ✔️TEST_PROFILE: all |
systemtest/src/test/java/io/strimzi/systemtest/security/oauth/OauthAuthorizationST.java
Outdated
Show resolved
Hide resolved
Signed-off-by: hzrncik <[email protected]>
|
/azp run regression |
|
Azure Pipelines successfully started running 1 pipeline(s). |
…afkas created in tests Signed-off-by: hzrncik <[email protected]>
henryZrncik
added
ready for merge
|
/azp run kraft-regression |
|
Azure Pipelines successfully started running 1 pipeline(s). |
Type of change
Description
stop using ephemeral kafka with 1 replica and use either persistent storage or more replicas.
Fixes #9748