-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
ACM 2.13.0
-
False
-
None
-
False
-
-
-
Low
-
None
Description of problem:
In the volsync test automation, we hit an issue with rsyncTLS, but seems like this issue can most likely happen with rsync as well. A fix could be made to address both at once.
The test failure is here: https://jenkins-csb-rhacm-tests.dno.corp.redhat.com/job/qe-acm-automation-poc/job/volsync_e2e_test_execution/463/
Essentially the problem comes down to this:
- The automation creates a replicationdestination
- The automation does things like copy the ssk key secret from the dest ns to the source (this is good, works just fine)
- The automation creates a replicationsource
I think what is happening is that at this point the replicationdestination does not have an address set yet (in status.rsyncTLS.address), so this code: https://gitlab.cee.redhat.com/rhacm-auto/e2e-backup-restore/-/blob/main/cypress/apis/volsync.js?ref_type=heads#L73
ends up not adding an address to the replicationsource spec.
- The replicationsource then fails as it does not have an address to connect to.
I checked on the system and the replicationdestination does have the address set in the status, so it's most likely a timing issue that the service did not populate an address in time.
I think we just need to add some sort of validation where we wait for the replicationdestination address to be populated in the status before we proceed and create the replicationsource.
I can help do this, but it's probably easiest for Thuy to make the change when she comes back.
Version-Release number of selected component (if applicable):
How reproducible:
Steps to Reproduce:
- ...