We’re trying to incorporate incremental backups in our couchbase cluster and I noticed that if the cbbackupmgr process fails (e.g. creating a repo that already exists), it does not delete the lock that it creates (lock.lk), causing subsequent calls to cbbackupmgr to fail. We’re using couchbase 6.5. It’s a problem for us since we are using cron jobs to schedule the backup regularly and lock failures causes the backups to stop. Is this a known problem? If not, what are your suggestions to get around that?
./cbbackupmgr config --archive ~/backup --repo users
Backup repository creation failed: Backup Repository `users` exists
lock.lk logs user
Hello @Jim_Quitevis and welcome to the Couchbase Forums.
There are cases where the lock files is not deleted, generally this does not cause problems.
cbbackupmgr will automatically remove the lock file the next time it runs by checking if the process that created the lock does not exist anymore.
I have tried to reproduce the problem but I could not. These are the steps I used on Mac OS with
Couchbase Server 6.5.0.
- Create the repo user:
[patrick:~] $ cbbackupmgr config --archive ~/backup --repo user
Backup repository `user` created successfully in archive `/Users/patrick/backup`
- Create the repo user again:
[patrick:~] $ couchbase-core/bin/cbbackupmgr config --archive ~/backup --repo user
Backup repository creation failed: Backup Repository `user` exists
- Create a new repo called users
[patrick:~] $ cbbackupmgr config --archive ~/backup --repo users
Backup repository `users` created successfully in archive `/Users/patrick/backup`
To help debug this further can can the following be provided:
- cbbackupmgr version
- Operating systems
- Is Kubernetes being used, what is the setup?
- Steps to reproduce (if possible)