Description:
I encountered this error when trying to run a backup when one pod in the innodb cluster was recovering from a restart.
```
Traceback (most recent call last):
File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 411, in command_do_create_backup
info = do_backup(backup, job_name, start, backup_dir, logger)
File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 321, in do_backup
backup_source = pick_source_instance(cluster, backup, profile, logger)
File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 265, in pick_source_instance
tmp = dba.get_cluster().status({"extended": 1})["defaultReplicaSet"]
~~~~~~~~~~~~~~~^^
RuntimeError: This function is not available through a session to an instance belonging to an unmanaged replication group
[ERROR] [backup] Backup failed with an exception: This function is not available through a session to an instance belonging to an unmanaged replication group
[INFO] [backup] Command execute-backup finished with code False
```
I believe this is due to the backup script selecting the pod that was not fully joined to the cluster yet. Once that pod joined the cluster, I was able to run backups successfully.
How to repeat:
Have 3 pod innodb cluster, restart one of the pods such that it comes back as ready, but not yet passing all readiness gates. Kick off a backup job
Suggested fix:
https://github.com/mysql/mysql-operator/blob/3b14f90575d2df410548176e1fc6acf7c7f2ba17/mysq...
The code that attempts to get the cluster status from the instance should either handle the "RuntimeError: This function is not available through a session to an instance belonging to an unmanaged replication group" error, or we should check if the pod is Ready and passing all readiness checks (https://github.com/mysql/mysql-operator/blob/3b14f90575d2df410548176e1fc6acf7c7f2ba17/mysq...) before attempting to connect to it
© 2026, Oracle Corporation and/or its affiliates
Description: I encountered this error when trying to run a backup when one pod in the innodb cluster was recovering from a restart. ``` Traceback (most recent call last): File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 411, in command_do_create_backup info = do_backup(backup, job_name, start, backup_dir, logger) File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 321, in do_backup backup_source = pick_source_instance(cluster, backup, profile, logger) File "/usr/lib/mysqlsh/python-packages/mysqloperator/backup_main.py", line 265, in pick_source_instance tmp = dba.get_cluster().status({"extended": 1})["defaultReplicaSet"] ~~~~~~~~~~~~~~~^^ RuntimeError: This function is not available through a session to an instance belonging to an unmanaged replication group [ERROR] [backup] Backup failed with an exception: This function is not available through a session to an instance belonging to an unmanaged replication group [INFO] [backup] Command execute-backup finished with code False ``` I believe this is due to the backup script selecting the pod that was not fully joined to the cluster yet. Once that pod joined the cluster, I was able to run backups successfully. How to repeat: Have 3 pod innodb cluster, restart one of the pods such that it comes back as ready, but not yet passing all readiness gates. Kick off a backup job Suggested fix: https://github.com/mysql/mysql-operator/blob/3b14f90575d2df410548176e1fc6acf7c7f2ba17/mysq... The code that attempts to get the cluster status from the instance should either handle the "RuntimeError: This function is not available through a session to an instance belonging to an unmanaged replication group" error, or we should check if the pod is Ready and passing all readiness checks (https://github.com/mysql/mysql-operator/blob/3b14f90575d2df410548176e1fc6acf7c7f2ba17/mysq...) before attempting to connect to it