Comment #13 on issue 944 by [email protected]: gnt-instance
replace-disks -s / -a fails if vm is running and secondary lvm-vol is gone
https://code.google.com/p/ganeti/issues/detail?id=944
Hi,
we've hit the same bug, and I'm afraid the above patch doesn't really work.
We have a 2 node cluster running Wheezy and Ganeti2.12 and we've upgraded
to kernel 3.16 with DRBD 8.4 from backports.
After doing resetup on the second node (meaning all lvs were wiped)
doing "replace-disks -s <vm>" failed with the same error as this ticket.
After commenting the Set_minor_SyncParams code:
ena.test.gnt.grnet.gr:~# gnt-instance replace-disks -s alexk.grnet.gr
Fri Nov 6 14:36:13 2015 Replacing disk(s) 0 for instance 'alexk.grnet.gr'
Fri Nov 6 14:36:13 2015 Current primary node: ena.test.gnt.grnet.gr
Fri Nov 6 14:36:13 2015 Current seconary node: dio.test.gnt.grnet.gr
Fri Nov 6 14:36:13 2015 - WARNING: Could not prepare block device disk/0
on node dio.test.gnt.grnet.gr (is_primary=False, pass=1): Error while
assembling disk: drbd0: can't setup network: exited with exit code 10 -
resource0: Failure: (158) Unknown resource\nadditional info from
kernel:\nunknown resource\n
Fri Nov 6 14:36:13 2015 STEP 1/6 Check device existence
Fri Nov 6 14:36:13 2015 - INFO: Checking disk/0 on ena.test.gnt.grnet.gr
Fri Nov 6 14:36:13 2015 - INFO: Checking disk/0 on dio.test.gnt.grnet.gr
Fri Nov 6 14:36:14 2015 - INFO: Checking disk/0 on ena.test.gnt.grnet.gr
Fri Nov 6 14:36:14 2015 - INFO: Checking disk/0 on dio.test.gnt.grnet.gr
Failure: command execution error:
Can't find disk/0 on node dio.test.gnt.grnet.gr: disk not found
Disks seem to be not properly activated. Try running activate-disks on the
instance before using replace-disks.
From the node-daemon.log:
2015-11-06 14:47:41,396: ganeti-noded pid=3346 INFO RunCmd drbdsetup
connect resource0 ipv4:<ip_local>:11001 ipv4:<ip_remote>:11001 --protocol C
--after-sb-0pri discard-zero-changes --after-sb-1pri consensus
--cram-hmac-alg md5 --shared-secret <..>
2015-11-06 14:47:41,400: ganeti-noded pid=3346 ERROR drbd0: can't setup
network: exited with exit code 10 - resource0: Failure: (158) Unknown
resource
additional info from kernel:
unknown resource
My understanding is that Ganeti doesn't initialize the resource without the
syncer command so it still fails.
When i downgraded the node to 8.3, replace-disks worked ok.
--
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
