Details
-
Bug
-
Resolution: Fixed
-
Major
-
None
-
Pre-Christmas Completions
-
1
Description
Problem
When an incremental backup is scheduled through the operator resource as the first backup, prior to a full backup occurring, it fails due to the backup repository only being created on a full backup.
This results in logic errors stating that the directory is 'Not an archive directory'.
The error in full can be seen below:
2023-11-14T15:53:01 INFO Timestamp: 2023-11-14 15:53:01.132742 |
2023-11-14T15:53:01 INFO Arguments: cluster=cb-example, mode=backup, full=False, incremental=True, backup_ret=720.0, disable_bucket_config=False, repo=None, start=None, end=None, map_data=None, filter_keys=None, filter_values=None, enable_bucket_config=False, force_updates=False, include_data=None, exclude_data=None, disable_views=False, disable_gsi_indexes=False, disable_ft_indexes=False, disable_ft_alias=False, disable_data=False, disable_analytics=False, disable_eventing=False, disable_cluster_analytics=False, disable_bucket_query=False, disable_cluster_query=False, cacert=None, log_ret=168.0, verbosity=INFO, s3_bucket=None, obj_store=None, obj_auth_by_instance_metadata=None, obj_endpoint=None, obj_cacert=None, s3_force_path_style=True, threads=1, default_recovery=none |
2023-11-14T15:53:01 INFO Checking connection to Kubernetes API... |
2023-11-14T15:53:01 ERROR Subprocess call exited with non-zero return code 1 |
2023-11-14T15:53:01 ERROR Arguments: cbbackupmgr info --json --archive /data/backups |
2023-11-14T15:53:01 ERROR Stdout: b'{"error":"Error opening archive at /data/backups due to `Not an archive directory`"}\n' |
2023-11-14T15:53:01 ERROR Stderr: b'' |
2023-11-14T15:53:01 ERROR Failed to list repositories |
2023-11-14T15:53:01 ERROR Command '['cbbackupmgr', 'info', '--json', '--archive', '/data/backups']' returned non-zero exit status 1. |
Steps to reproduce
- Deploy Couchbase Autonomous Operator in a Kubernetes environment by following the CAO install documentation.
- Define a configuration file that contains the backup resource. An example configuration file (testCluster.yaml) is attached.
- Create a Couchbase Cluster via the configuration file and schedule the backup as a cronJob.
- Wait for the backup to occur and if the incremental backup is scheduled prior to a full backup, this error will occur and the backup will not be successful.
Expectations
It is expected that the repository is created on the first occurrence of a backup, regardless of the type being an incremental backup.
Attachments
Issue Links
- relates to
-
K8S-3354 Cloud Native 2.6.1 - Release Notes
- Closed