[DRBD-user] Weird HA-behavior with XEN/HA/DRBD/LVM

sussox sussox at gmail.com
Tue Jun 30 15:01:32 CEST 2009

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


I have a HA-setup with Xen heartbeat lvm and DRBD. According to this howto:
http://www.asplund.nu/xencluster/xen-cluster-howto.html

Im using 2 Poweredge R710. ha1 & ha2 with Ubuntu 8.04

I can do a manual live migration of the domU's without any problems. Also if
i shutdown heartbeat on ha1 with init.d/heartbeat the domU's are migrated to
ha2 sucessfully.

Also, if i pull the plug on ha1, after a while the domU's start on ha2 (as
they should).

However! When doing a "reboot" on ha1, domU's begin to migrate but then
crashes on ha2. pasting ha-debug and xend.log below.

Any ides why it keeps doing this? All i can think of is that some process is
being killed to fast (when the domU's are beeing migrated, but i don't know
what to look for.) Also, i ran a test couple of weeks ago with the same
setup but on one R710 and a older shuttle and then there was no problem.
Tried to redo the howto twice but with the same problem

Cheers! /Sussox

ha-debug:

Code:

heartbeat[28295]: 2009/06/30_13:46:12 info: Received shutdown notice from
'ha1.vbm.se'.
heartbeat[28295]: 2009/06/30_13:46:12 info: Resources being acquired from
ha1.vbm.se.
heartbeat[28295]: 2009/06/30_13:46:12 debug: StartNextRemoteRscReq(): child
count 1
heartbeat[29445]: 2009/06/30_13:46:12 info: acquire local HA resources
(standby).
ResourceManager[29472]:	2009/06/30_13:46:12 info: Acquiring resource group:
ha2.vbm.se xendomainsHA2
heartbeat[29446]: 2009/06/30_13:46:13 info: Local Resource acquisition
completed.
heartbeat[28295]: 2009/06/30_13:46:13 debug: StartNextRemoteRscReq(): child
count 2
heartbeat[28295]: 2009/06/30_13:46:13 debug: StartNextRemoteRscReq(): child
count 1
ResourceManager[29472]:	2009/06/30_13:46:13 info: Running
/etc/ha.d/resource.d/xendomainsHA2  start
ResourceManager[29472]:	2009/06/30_13:46:13 debug: Starting
/etc/ha.d/resource.d/xendomainsHA2  start
ResourceManager[29472]:	2009/06/30_13:46:13 debug:
/etc/ha.d/resource.d/xendomainsHA2  start done. RC=0
heartbeat[29445]: 2009/06/30_13:46:13 info: local HA resource acquisition
completed (standby).
heartbeat[28295]: 2009/06/30_13:46:13 info: Standby resource acquisition
done [foreign].
heartbeat[29559]: 2009/06/30_13:46:13 debug: notify_world: setting SIGCHLD
Handler to SIG_DFL
harc[29559]:	2009/06/30_13:46:13 info: Running /etc/ha.d/rc.d/status status
mach_down[29573]:	2009/06/30_13:46:13 info: Taking over resource group
xendomainsHA1
ResourceManager[29597]:	2009/06/30_13:46:13 info: Acquiring resource group:
ha1.vbm.se xendomainsHA1
ResourceManager[29597]:	2009/06/30_13:46:13 info: Running
/etc/ha.d/resource.d/xendomainsHA1  start
ResourceManager[29597]:	2009/06/30_13:46:13 debug: Starting
/etc/ha.d/resource.d/xendomainsHA1  start
Starting auto Xen domains: hejsan(skip) *   [done] 
ResourceManager[29597]:	2009/06/30_13:46:13 debug:
/etc/ha.d/resource.d/xendomainsHA1  start done. RC=0
mach_down[29573]:	2009/06/30_13:46:13 info: /usr/share/heartbeat/mach_down:
nice_failback: foreign resources acquired
mach_down[29573]:	2009/06/30_13:46:13 info: mach_down takeover complete for
node ha1.vbm.se.
heartbeat[28295]: 2009/06/30_13:46:13 info: mach_down takeover complete.
heartbeat[29696]: 2009/06/30_13:46:13 debug: notify_world: setting SIGCHLD
Handler to SIG_DFL
harc[29696]:	2009/06/30_13:46:13 info: Running
/etc/ha.d/rc.d/ip-request-resp ip-request-resp
ip-request-resp[29696]:	2009/06/30_13:46:13 received ip-request-resp
xendomainsHA2 OK yes
ResourceManager[29715]:	2009/06/30_13:46:13 info: Acquiring resource group:
ha2.vbm.se xendomainsHA2
ResourceManager[29715]:	2009/06/30_13:46:13 info: Running
/etc/ha.d/resource.d/xendomainsHA2  start
ResourceManager[29715]:	2009/06/30_13:46:13 debug: Starting
/etc/ha.d/resource.d/xendomainsHA2  start
ResourceManager[29715]:	2009/06/30_13:46:13 debug:
/etc/ha.d/resource.d/xendomainsHA2  start done. RC=0
heartbeat[28295]: 2009/06/30_13:46:24 WARN: node ha1.vbm.se: is dead
heartbeat[28295]: 2009/06/30_13:46:24 info: Dead node ha1.vbm.se gave up
resources.
heartbeat[28295]: 2009/06/30_13:46:24 info: Link ha1.vbm.se:eth0 dead.

xend.log

Code:

[2009-06-30 13:46:11 5499] DEBUG (XendCheckpoint:210) restore:shadow=0x0,
_static_max=0x18000000, _static_min=0x0, 
[2009-06-30 13:46:11 5499] DEBUG (balloon:151) Balloon: 398436 KiB free;
need 393216; done.
[2009-06-30 13:46:11 5499] DEBUG (XendCheckpoint:227) [xc_restore]:
/usr/lib/xen/bin/xc_restore 4 7 1 2 0 0 0
[2009-06-30 13:46:11 5499] INFO (XendCheckpoint:365) xc_domain_restore
start: p2m_size = 18800
[2009-06-30 13:46:11 5499] INFO (XendCheckpoint:365) Reloading memory pages:  
0%
[2009-06-30 13:46:14 5499] INFO (XendCheckpoint:365) ERROR Internal error:
Error when reading page (type was 0)
[2009-06-30 13:46:14 5499] INFO (XendCheckpoint:365) Restore exit with rc=1
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1913)
XendDomainInfo.destroy: domid=7
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1930)
XendDomainInfo.destroyDomain(7)
[2009-06-30 13:46:14 5499] ERROR (XendDomainInfo:1942)
XendDomainInfo.destroy: xc.domain_destroy failed.
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/xen/xend/XendDomainInfo.py", line
1937, in destroyDomain
    xc.domain_destroy(self.domid)
Error: (3, 'No such process')
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1553) No device model
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1555) Releasing devices
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1561) Removing vif/0
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:590)
XendDomainInfo.destroyDevice: deviceClass = vif, device = vif/0
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1561) Removing vbd/51713
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:590)
XendDomainInfo.destroyDevice: deviceClass = vbd, device = vbd/51713
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1561) Removing vbd/51714
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:590)
XendDomainInfo.destroyDevice: deviceClass = vbd, device = vbd/51714
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:1561) Removing console/0
[2009-06-30 13:46:14 5499] DEBUG (XendDomainInfo:590)
XendDomainInfo.destroyDevice: deviceClass = console, device = console/0
[2009-06-30 13:46:14 5499] ERROR (XendDomain:1136) Restore failed
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/xen/xend/XendDomain.py", line 1134,
in domain_restore_fd
    return XendCheckpoint.restore(self, fd, paused=paused)
  File "/usr/lib/python2.5/site-packages/xen/xend/XendCheckpoint.py", line
231, in restore
    forkHelper(cmd, fd, handler.handler, True)
  File "/usr/lib/python2.5/site-packages/xen/xend/XendCheckpoint.py", line
353, in forkHelper
    raise XendError("%s failed" % string.join(cmd))
XendError: /usr/lib/xen/bin/xc_restore 4 7 1 2 0 0 0 failed


-- 
View this message in context: http://www.nabble.com/Weird-HA-behavior-with-XEN-HA-DRBD-LVM-tp24271462p24271462.html
Sent from the DRBD - User mailing list archive at Nabble.com.




More information about the drbd-user mailing list