Device already in use

[willifehler]

Hey @lae,

if I restart my Playbook the deployment will fail. Maybe a check would be helpful if there is already an existing OSD?

8961 failed: [...] (item={'device': '/dev/disk/by-id/nvme-MO003200KYDNC_S70NNT0X605953'}) => {"ansible_loop_var": "item", "changed": true, "cmd": ["pveceph", "osd", "create", "/dev/disk/by-id/nvme-MO003200KYDNC_S70NNT0X605953"], "delta": "0:00:00.404626", "end": "2024-09-12 07:24:01.340003", "item": {"device": "/dev/disk/by
 8962 -id/nvme-MO003200KYDNC_S70NNT0X605953"}, "msg": "non-zero return code", "rc": 25, "start": "2024-09-12 07:24:00.935377", "stderr": "device '/dev/nvme3n1' is already in use", "stderr_lines": ["device '/dev/nvme3n1' is a

Cheers - Willi

[lae]

My understanding is that if the role provisioned the OSDs before, then it shouldn't attempt to reprovision them again.

• your output is truncated
• post your role variables
• post the whole task

Maybe a check would be helpful

https://github.com/lae/ansible-role-proxmox/blob/develop/tasks/ceph.yml#L45

The logic for what you think is the solution is there. The context for your issue/symptoms is not.

maintainer side comment:
I think it may be more productive for you to start your own discussion thread for your experience preparing/administering a playbook using this role, and then opening issues with actionable information when relevant.

Please note that I am not the only contributor to this role. Mentioning someone unsolicited on a bug report for a part of the role they didn't implement and isn't super familiar with is...unsettling.