Description:
Trying to use initDB object i deleted the object:
kubectl -n mysql delete InnoDBCluster mysql
innodbcluster.mysql.oracle.com "mysql" deleted
but this action got stuck.
In operator logs i can only see this printed in loop:
[2023-07-19 14:42:08,848] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules
[2023-07-19 14:42:08,852] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running
[2023-07-19 14:42:18,997] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules
[2023-07-19 14:42:19,000] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running
[2023-07-19 14:42:29,140] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules
[2023-07-19 14:42:29,144] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running
InnoDBCluster object status
kubectl -n mysql get InnoDBCluster
NAME STATUS ONLINE INSTANCES ROUTERS AGE
mysql INITIALIZING 0 3 2 44m
Only operator pod is running:
mysql-operator-79d96c78c9-xn9qk 1/1 Running 0 21d
Now i m stuck, as i can't delete the InnoDBCluster and if i run helm upgrade nothing happens.
Thanks,
Carlos
How to repeat:
I have:
1-deploy cluster
2-added initDB values:
initDB:
dump:
name:
path: mysql-backup-schedule230719030102
# options:
# includeSchemas:
persistentVolumeClaim:
claimName: backup-volume-claim-1
3- I did helm upgrade to create InnoDBCluster again - Restore didn't worked
4- Deleted InnoDBCluster to see if create InnoDBCluster again would work
5- helm upgrade to create the InnoDBCluster object - Restore didn't worked
6- after troubleshooting the helm templates realised that name was needed (missing the documentation)
7- Deleted InnoDBCluster to see if create InnoDBCluster again would work
After this second delete got to that state.
Suggested fix:
Deleting a InnoDBCluster object should not get stuck like this, as explained now i can't delete or create the cluster again.
Also in https://dev.mysql.com/doc/mysql-operator/en/mysql-operator-properties.html#mysql-operator-... is state that name is not required (false) but seems wrong, looking to helm template:
..
{{- if and .name (or .ociObjectStorage .s3 .persistentVolumeClaim .options) }}
initDB:
dump:
..
it is needed, docs should be fixed required to true
Description: Trying to use initDB object i deleted the object: kubectl -n mysql delete InnoDBCluster mysql innodbcluster.mysql.oracle.com "mysql" deleted but this action got stuck. In operator logs i can only see this printed in loop: [2023-07-19 14:42:08,848] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules [2023-07-19 14:42:08,852] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running [2023-07-19 14:42:18,997] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules [2023-07-19 14:42:19,000] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running [2023-07-19 14:42:29,140] kopf.objects [INFO ] on_innodbcluster_field_backup_schedules [2023-07-19 14:42:29,144] kopf.objects [ERROR ] Handler 'on_innodbcluster_field_backup_schedules/spec.backupSchedules' failed temporarily: The cluster is not ready. Will create the schedules once the first instance is up and running InnoDBCluster object status kubectl -n mysql get InnoDBCluster NAME STATUS ONLINE INSTANCES ROUTERS AGE mysql INITIALIZING 0 3 2 44m Only operator pod is running: mysql-operator-79d96c78c9-xn9qk 1/1 Running 0 21d Now i m stuck, as i can't delete the InnoDBCluster and if i run helm upgrade nothing happens. Thanks, Carlos How to repeat: I have: 1-deploy cluster 2-added initDB values: initDB: dump: name: path: mysql-backup-schedule230719030102 # options: # includeSchemas: persistentVolumeClaim: claimName: backup-volume-claim-1 3- I did helm upgrade to create InnoDBCluster again - Restore didn't worked 4- Deleted InnoDBCluster to see if create InnoDBCluster again would work 5- helm upgrade to create the InnoDBCluster object - Restore didn't worked 6- after troubleshooting the helm templates realised that name was needed (missing the documentation) 7- Deleted InnoDBCluster to see if create InnoDBCluster again would work After this second delete got to that state. Suggested fix: Deleting a InnoDBCluster object should not get stuck like this, as explained now i can't delete or create the cluster again. Also in https://dev.mysql.com/doc/mysql-operator/en/mysql-operator-properties.html#mysql-operator-... is state that name is not required (false) but seems wrong, looking to helm template: .. {{- if and .name (or .ociObjectStorage .s3 .persistentVolumeClaim .options) }} initDB: dump: .. it is needed, docs should be fixed required to true