Cannot backup to S3 provider (Backblaze) due to missing S3 options

[ToshY]

Report

Problem

Both the s3 options here, as well as in the Helm chart for pxc-db are outdated. There are S3 options needed in order to have working backups for other (semi-) S3-compatible services.

Since AWS 1.37.0 / 2.23.0, integrity protection was enabled by default: aws/aws-cli#9214 This causes issues for providers like Backblaze that do not support headers like x-amz-checksum-algorithm, leading to errors:

2025/04/19 16:49:21 ERROR: collect binlog files: manage binlog: put binlog.000003 object: put object binlog_1745071805_3037c43d8f1b1bd390e40dba73d07963: Unsupported header 'x-amz-checksum-algorithm' received for this API call.

More about the problem

Options that are missing and causing issues:

• forcePathStyle - bucketname part of path / PSMDB-1492-feature-forcePathStyle percona-helm-charts#349
• prefix - specific directory in bucket / Add support for s3 backup Prefix percona-helm-charts#322 / Backups can't be stored to a set prefix in a storage location #1991
• checksumAlgorithm - need to be disabled/empty for Backblaze / https://www.backblaze.com/docs/cloud-storage-s3-compatible-api#unsupported-features

The implementation of S3 backup options seem very inconsistent across different operators. I first read about backup/restore on a blog post here for MongoDB:
https://www.percona.com/blog/percona-operator-for-mongodb-backup-and-restore-on-s3-compatible-storage-backblaze/

If the options that are used in the percona/percona-server-mongodb-operator would be implemented here, it would already go a long way: https://github.com/percona/percona-server-mongodb-operator/blob/676b4db0183dc4721ac10ce75cb6981802e78aca/deploy/crd.yaml#L124

Steps to reproduce

backup:
  enabled: true
  image:
    repository: percona/percona-xtradb-cluster-operator
  pitr:
    enabled: true
    storageName: s3-eu-central
    timeBetweenUploads: 300
    timeoutSeconds: 60
  storages:
    s3-eu-central:
      type: s3
      verifyTLS: true
      s3:
        bucket: <bucket-name>
        credentialsSecret: <secret-name>
        region: 'eu-central-003'
        endpointUrl: 'https://s3.eu-central-003.backblazeb2.com'

Versions

1. Kubernetes - k3s (v1.31.7)
2. Operator - percona/percona-xtradb-cluster-operator ; helm chart 1.17.0
3. Database - PXC

Anything else?

p.s. This is the first time using Percona software

[ToshY]

Can I get feedback from one of the maintainers? Is the issue missing information or am I missing something regarding the s3 configuration?

[ToshY]

Could I please get some feedback? I do not know if this will be picked up (eventually) or not.

Will Percona look into creating a more robust S3 backup implementation that can work for multiple S3-compatible providers (consistent across different Percona operators)?

With that information I can make the informed decision to either wait or start looking for alternatives.