Error: Acquire running lock failed: 256 with HA after upgr. to 2.1.5

Question

Hello,
after upgrading the 2 servers in server pool from 2.1.2 to 2.1.5 using yum, all vms throw:
Error: Acquire running lock failed: 256
when trying to start them. HA-infrastructure-check works well from OVMM 2.1.2 (not upgraded yet) and if we disable HA for the server pool, the vms come up. Shared storage is on NFS
We have the following in /var/log/xen/xend.log:
[2009-06-26 14:52:21 2649] DEBUG (XendDomainInfo:87) XendDomainInfo.create(['vm', ['name', '470_vm215'], ['memory', '1024'], ['on_reboot', 'restart'], ['on_crash', 'restart'], ['vcpus', 1], ['on_xend_start', 'ignore'], ['on_xend_stop', 'ignore'], ['uuid', '14205f6a-12cc-42da-a32e-06f30b6288d4'], ['bootloader', '/usr/bin/pygrub'], ['bootloader_args', '-q'], ['image', ['linux']], ['device', ['vbd', ['uname', 'file:/OVS/running_pool/470_vm215/system.img'], ['dev', 'sda'], ['mode', 'w']]], ['device', ['vif', ['bridge', 'xenbr0'], ['mac', '00:16:3E:32:CD:FC'], ['type', 'netfront']]], ['device', ['vkbd']], ['device', ['vfb', ['vncunused', '1'], ['vncpasswd', 'XXXXXXXX'], ['vnclisten', '0.0.0.0'], ['type', 'vnc'], ['display', 'localhost:11.0'], ['xauthority', '/root/.Xauthority']]]])
[2009-06-26 14:52:21 2649] ERROR (XendDomainInfo:452) VM start failed
Traceback (most recent call last):
File "/usr/lib/python2.4/site-packages/xen/xend/XendDomainInfo.py", line 430, in start
self.acquire_running_lock();
File "/usr/lib/python2.4/site-packages/xen/xend/XendDomainInfo.py", line 2891, in acquire_running_lock
raise XendError('Acquire running lock failed: %s' % status)
XendError: Acquire running lock failed: 256
[2009-06-26 14:52:21 2649] DEBUG (XendDomainInfo:1881) XendDomainInfo.destroy: domid=None
[2009-06-26 14:52:21 2649] ERROR (XendDomainInfo:92) Domain construction failed
Traceback (most recent call last):
File "/usr/lib/python2.4/site-packages/xen/xend/XendDomainInfo.py", line 90, in create
vm.start()
File "/usr/lib/python2.4/site-packages/xen/xend/XendDomainInfo.py", line 430, in start
self.acquire_running_lock();
File "/usr/lib/python2.4/site-packages/xen/xend/XendDomainInfo.py", line 2891, in acquire_running_lock
raise XendError('Acquire running lock failed: %s' % status)
XendError: Acquire running lock failed: 256
[2009-06-26 14:52:21 2649] DEBUG (XendDomainInfo:1881) XendDomainInfo.destroy: domid=None
and in /var/log/xen/xend-debug.log:
ha_set_dlm_lock ['14205f6a-12cc-42da-a32e-06f30b6288d4', '470_vm215'] =>
failed: <Exception: create lock('/dlm/ovm/4205f6a12cc42daa32e06f30b6288d4') failed. <OSError: [Errno 26] Text file busy: '/dlm/ovm/4205f6a12cc42daa32e06f30b6288d4'>
StackTrace:
File "/opt/ovs-agent-2.2/OVSXHA.py", line 371, in ha_set_dlm_lock
fd = os.open(lock_path, os.O_CREAT | os.O_NONBLOCK)
>
StackTrace:
File "/opt/ovs-agent-2.2/OVSXHA.py", line 379, in ha_set_dlm_lock
raise Exception("create lock('%s') failed. %s" % (lock_path, errmsg))
Any ideas?
Thanks, Olaf

Answer 1

suk wrote:
1. To delete the database, which database are you referring to? I'm referring to the ovs-agent database:
1. Stop the agent: # /etc/init.d/ovs-agent stop 2. Delete the database: # rm -f /etc/ovs-agent/db 3. Start the agent: # /etc/init.d/ovs-agent start If you have Oracle VM Manager v2.1.5 (which you should have upgraded before your servers, btw), you can now try using the "Restore" option for the server pool. Click on the Server Pool in the manager and hit "Restore" -- this will send the configuration data back down to the agents in the pool. Alternatively, you may need to de-register the servers in your pool and re-register them.
2. To re-imported the guests, how?In Oracle VM Manager -- if you recreate the pool, you have the option of "Importing" both VM Templates and VM Guests. Deleting the pool in the Manager does not delete your guests as long as you ensure that you DO NOT select "Remove Working Directories".