Status: New
Owner: ----
New issue 606 by [email protected]: replace-disks report incorrect error
message and do not fix when offline
http://code.google.com/p/ganeti/issues/detail?id=606
-> Ganeti version: 2.8.0
-> What distribution are you using?
Debian 7
-> What steps will reproduce the problem?
1. Create a two node cluster
2. delete/recreate an lv or the vg in the secondary
3. run repair-disks with the
This is a two-nodes cluster and I had to recreate from scratch the volume
group on the secondary due to an update on the RAID5 volume.
-> What is the expected output? What do you see instead?
I tried to run "replace-disks" both with automatic and forcing the
secondary, but these are the output:
====================================
# gnt-instance replace-disks -s fw-melano
Fri Oct 18 19:33:18 2013 Replacing disk(s) 0 for instance
'fw-melano.wiran.net'
Fri Oct 18 19:33:18 2013 Current primary node: lugspa0638.XX.XX
Fri Oct 18 19:33:18 2013 Current seconary node: lugspb0638.XX.XX
Fri Oct 18 19:33:18 2013 STEP 1/6 Check device existence
Fri Oct 18 19:33:18 2013 - INFO: Checking disk/0 on lugspa0638.XX.XX
Fri Oct 18 19:33:19 2013 - INFO: Checking disk/0 on lugspb0638.XX.XX
Failure: command execution error:
Can't find disk/0 on node lugspb0638.moresi.com: disk not found
====================================
# gnt-instance replace-disks --auto fw-melano
Fri Oct 18 19:34:01 2013 - INFO: Checking disk/0 on lugspa0638.XX.XX
Fri Oct 18 19:34:02 2013 - INFO: Checking disk/0 on lugspb0638.XX.XX
Failure: prerequisites not met for this operation:
error type: wrong_state, error details:
Please run activate-disks on instance fw-melano.wiran.net first
====================================
The log on the secondary:
====================================
2013-10-18 19:34:33,790: ganeti-noded pid=18330 INFO RunCmd lvs
--noheadings --separator=, --units=k --nosuffix
-olv_attr,lv_kernel_major,lv_kernel_minor,vg_extent_size,stripes
/dev/virtvg/91e9815c-c83d-4888-b2f6-55dddf301226.disk0_meta
2013-10-18 19:34:33,815: ganeti-noded pid=18330 ERROR Can't find LV
/dev/virtvg/91e9815c-c83d-4888-b2f6-55dddf301226.disk0_meta: exited with
exit code 5, One or more specified logical volume(s) not found.
====================================
Well, it's obvious that the lv is not found ... I deleted the entire vg
-> Please provide any additional information below.
The commands should (with these prority)
a) In the short term, report that the guest is not off-line instead of just
failing
b) In the longer term, i.e. next release, allow on-line rebuild of
secondaries... it could be a problem with running services.
Thanks a lot
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
