Garbage collection fails with Google Cloud Bucket

[rmundel]

Hey guys, I started to using the new S3 Endpoint for sync our local backups.

The backup process went fine with one 800ish GB snapshot, but the GC gave me the following error in the task log:

2025-10-03T00:00:00-04:00: starting garbage collection on store google-cloud
2025-10-03T00:00:00-04:00: task triggered by schedule 'daily'
2025-10-03T00:00:01-04:00: Access time update check successful, proceeding with GC.
2025-10-03T00:00:01-04:00: Using access time cutoff 1d 5m, minimum access time is 2025-10-02T03:55:00Z
2025-10-03T00:00:01-04:00: Start GC phase1 (mark used chunks)
2025-10-03T00:00:02-04:00: marked 50% (1 of 2 index files)
2025-10-03T00:00:02-04:00: marked 100% (2 of 2 index files)
2025-10-03T00:00:02-04:00: Start GC phase2 (sweep unused chunks)
2025-10-03T00:00:02-04:00: queued notification (id=1c695637-0675-443f-8476-ba2e305c242b)
2025-10-03T00:00:02-04:00: TASK ERROR: failed to list chunk in s3 object store: failed to parse response body: custom: missing field `StorageClass`

I'm using a Nearline class bucket.

Any thoughts on how to fix the GC?

 

[Chris]

Hi,
different storage classes are currently not supported by the PBS S3 implementation. Could you verify the issue persists also on the standard storage class? For available classes see https://cloud.google.com/storage/docs/storage-classes

The response should include the storage class according to https://cloud.google.com/distribute...ice/storage/storage-s3-rest-api#ListObjectsV2
Can you further verify that this is reproducible and not just a transient error?

 

[techdude45]

Just chiming in here with the same issue. Using the S3 endpoint for GCP on my setup results in the same missing field task error. I have tried recreating the endpoint, and clearing and recreating my backup configuration, but it is the same. I have been running various verification and pruning jobs to try to see if something will get me past this, but so far it's consistently happening.