Project

General

Profile

Bug #58157

Ceph Orchestrator fails to recognize partition.

Added by Brian Woods 2 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
4 - irritation
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

I have a system that has an unmounted XFS volume (it was failing and was unmounted) that was created using the root device (so /dev/sdk rather than /dev/sdk1), not a partition ID. That was done for reasons that no longer apply to the system and is being removed.

This causes the volume to show as available, but when the Ceph Orchestrator tries to deploy an OSD, it fails. Generating a message like this:
/usr/bin/docker: stderr RuntimeError: Device /dev/sdk has a filesystem

This sends Ceph into a bit of a deployment loop and it will constantly repeat this process.

I have since removed the drive and can no further testing on this, but should be easy enough to re-product in a test lab.

Also available in: Atom PDF