docs(backups): state what the MariaDB dump delete does not cover

The comment on who deletes a MariaDB Backup listed namespace teardown
without saying that teardown can remove the operator Backup first, in
which case the dump is left behind with a Warning, so it promised a
delete the code does not make. The remover contract claimed a missing
key is never an error, which holds only for a backend that answers the
delete with 204, the one status minio-go accepts.

Assisted-by: LLM
Signed-off-by: Alexey Artamonov <aleksei.artamonov@aenix.io>
This commit is contained in:
Alexey Artamonov committed 2026-11-11 15:09:15 +03:00
1 parent 3334152da7
commit d57ebbe808
2 files changed
+6 -2

No files matched your search

+3 -1
View File
@@ -39,7 +39,9 @@ type s3ObjectLister interface {
}
type s3ObjectRemover interface {
// RemoveObject deletes key; a key that does not exist is not an error.
// RemoveObject deletes key. S3 answers 204 for a key that does not
// exist and minio-go takes 204 as success; a backend answering a
// missing key with any other status makes it an error.
RemoveObject(ctx context.Context, bucket, key string) error
}
@@ -142,7 +142,9 @@ func (r *BackupReconciler) cleanupOnDelete(ctx context.Context, backup *backupsv
// a dump on the platform bucket is deleted here or never. A dump in
// a tenant's own bucket is deleted too: the Backup proves it owns
// it, and tenants cannot delete Backups, so the delete comes from an
// administrator or from namespace teardown (see cleanupMariaDBBackup).
// administrator or from namespace teardown. Teardown can remove the
// operator Backup first, and the dump is then left behind with a
// Warning, since its storage is read from it (see cleanupMariaDBBackup).
return r.cleanupMariaDBBackup(ctx, backup)
case strategyv1alpha1.MongoDBStrategyKind:
// Split by flow (see cleanupMongoDBBackup). A legacy backup writes to the