Project

General

Profile

Actions
Copy link

Bug #44463

open

prepare fails if devices are already prepared for the current cluster

Added by Jan Fajerski about 4 years ago. Updated about 4 years ago.

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

0%

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

Description

There is a concern about prepares behaviour when it find the passed devices are already prepared for the current cluster. There is some validity to succeeding when a call to prepare finds an already prepared OSD.

We need to figure out which criteria need to be fulfilled to consider this scenario a success. The already prepared osd belonging to the same cluster is an obvious one. Other not so obvious ones are

  • osd id is in crushmap
  • osd has the same layout (size, multiple devices, filestore/bluestore)
  • ...
Actions
Copy link
 #1

Updated by Jan Fajerski about 4 years ago

Maybe this needs a flag to use this new behaviour in order to be kind to existing users relying on the current behaviour.

Actions
Copy link

Also available in: Atom PDF