-
Notifications
You must be signed in to change notification settings - Fork 100
Open
Description
Describe the bug
The migration process is stucking on but, the job batch is completed!
To Reproduce
Execute a simple command of migration like this:
pv-migrate migrate --source-namespace xx-xxxxxx --dest-namespace xx-xxxxxx --ignore-mounted xxxx-xxxxxxxxxx-xxx-pv-claim xxxx-xxxxxxxxxx-xxx-pvc --log-level=debug
Expected behavior
The completed process migration without intervention needed on pod job batch
Console output
🚀 Starting migration
💡 PVC xxxx-xxxxxxxxxx-xxx-pv-claim is mounted to node gke-staging-cluster-2-pool-01-1b6099b0-2vq1, ignoring...
💭 Will attempt 3 strategies: mnt2, svc, lbsvc
🚁 Attempting strategy: mnt2
creating 2 resource(s)
beginning wait for 2 resources with timeout of 1m0s
🔶 Cannot tail logs to display progress
Version
- Source and destination Kubernetes versions:
v1.25.16 pv-migrateversion and architecture:v1.7.1 - arm64- Installation method:
homebrew - Source and destination PVC type, size and accessModes
ReadWriteMany, 1TB, NFS -> ReadWriteOnce, 20 GB, pd.csi.storage.gke.io
Additional context
Clearly the process is completed, however, the binary does not detect the real state of the migration process in the pod (job batch), as can be seen in the last lines of the pod log:
sent 1,884,871,431 bytes received 832,045 bytes 28,356,443.25 bytes/sec
total size is 2,484,844,053 speedup is 1.32
+ rc=0
+ break
+ '[' 0 -ne 0 ]
+ exit 0
The pod state:
pv-migrate-cadca-rsync-7vwhn 1/2 NotReady 0 19m
Is needed remove the pod for the migration process is completed!
Metadata
Metadata
Assignees
Labels
No labels