Project

General

Profile

Bug #20525

ceph osd replace problem with osd out

Added by peng zhang over 4 years ago. Updated over 4 years ago.

Status:
Need More Info
Priority:
Normal
Assignee:
-
Category:
-
Target version:
% Done:

0%

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

Description

i have try the new function of replace the osd with new command ,it work, but i have some problem,i don't know if it's a problem

these is my step:
1、stop osd.4
  1. systemctl stop ceph-osd@4
2、destroy osd.4
  1. ceph osd destroy osd.4 --yes-i-really-mean-it
3、with new osd uuid
  1. uuidgen
    4576d84d-7677-42b8-810c-2dc39c589b2a
    and now the osd.4 status is down in ,so no backfill no recovery
    then i run command:
  2. ceph osd new 4576d84d-7677-42b8-810c-2dc39c589b2a 4
    and then osd.4 status is down out ,and backfill or recovery begin

4、replace osd.4 with the new uuid above
#ceph-disk prepare --osd-uuid 4576d84d-7677-42b8-810c-2dc39c589b2a /dev/sdf

and all pg active+clean

my problem is:

ceph osd new command may not mark the osd out ,and when the new replce osd add in,just make recovery begin.

History

#1 Updated by Kefu Chai over 4 years ago

  • Assignee deleted (Kefu Chai)

#2 Updated by Kefu Chai over 4 years ago

peng,

i don't follow you. could you rephrase your problem? what is the expected behavior?

#3 Updated by Kefu Chai over 4 years ago

  • Status changed from New to Need More Info

Also available in: Atom PDF