As per PR #7281, if repository count is more than 1, then snapshots deletion is achieved with a fast way, then we should have more than 1 FS backup repository per backup.
Signed-off-by: danfengl <danfengl@vmware.com>
1. Add sleep to avoid snapshot limitation issue https://docs.aws.amazon.com/AWSEC2/latest/APIReference/errors-overview.html#:~:text=SnapshotCreationPerVolumeRateExceeded;
2. Move InstallVelero variable out of struct of Veleroconfig as a global one since it's not for controlling any individual case;
3. Unskip migration test case on AWS pipeline, because we added a new EKS pipeline and deleted TKG AWS pipline in internal E2E test, so this restriction for TKG AWS pipline is no long existed;
4. Skip retainPV test on vSphere pipeline due to PV longtime bounding issue;
5. Fix failing get snapshot by CSI from EC2 issue, snapshot by CSI has no label of backup name.
Signed-off-by: danfengl <danfengl@vmware.com>
1. Fix issue of kubectl client and server mismatch version in GitAction E2E job, refer to https://github.com/elastic/cloud-on-k8s/issues/4737;
2. Adapt to the changing of keyword for involing Kpoia as fs backupper, new installtion breaked upgrade and migration tests;
3. Accept multi-labels of Ginkgo focus as input of E2E make command;
4. Distinguish workload namespace from each tests;
5. Fix issues of not using Velero util to perform Velero commands;
6. Add snapshot test case for NamespaceMapping E2E test;
7. Collect debug bundle after catching error of Velero backup or restore command;
Signed-off-by: danfengl <danfengl@vmware.com>
1. Add checkpoint in snapshot E2E test to verify snapshot CR should be created and snapshot should be created in cloud side after backup completion;
2. Fix snapshot name issue that CSI snapshot name in cloud side is not the same with other non-CSI cloud snapshots;
Signed-off-by: danfengl <danfengl@vmware.com>
