This post is more than 5 years old
6 Posts
0
1110
November 17th, 2010 06:00
Autostart 5.2.2 'Path Failed' Error Starting Backbone Service
Hi - I am running a 2 node Autostart 5.2.2 SQL cluster, and I am having a problem with one of the nodes - I am unable to start the backbone service. This cluster was working fine for a couple of years, and I just notice a few weeks ago that one of the nodes wasn't running. After a little digging, I found the following in the backbone log file:
ID00004722 Backbone release V3.3.1 #106 Dec 7 2006 02:00:00
site is now coming up, site-id <1> log_dir <..\log\backbone>
detect site-failure after: 120 secs
failure detection parameters: ratio=1/2 min=50
Backbone info Tue Nov 02 15:44:49 2010
ID00004353 restart Backbone: test
Backbone info Tue Nov 02 15:44:49 2010
ID00004530 Initial path to site 2 using ip=128.127.84.48 pxi=0x326398 ints=1
Backbone info Tue Nov 02 15:45:06 2010
ID00004549 Recompute site 2/125's IP address to 128.127.84.48: Path Failed
Backbone info Tue Nov 02 15:45:14 2010
(This repeats about 10 more times, but I thought I'd keep this a little shorter)
ID00004549 Recompute site 2/125's IP address to 128.127.84.48: Path Failed
Backbone info Tue Nov 02 15:46:45 2010
ID00004365 : shutdown (termination of detected)
I don't see any issues on the server with the network connection - I can ping back and forth between the servers on either NIC - any ideas?
Thanks,
John
tribicic
157 Posts
0
November 17th, 2010 06:00
Is the mentioned IP address correct? If it is and you are certain that the network works both ways, then you might want to try restarting the Autostart service on the active node. Just recently I had a similar situation and after restarting the Autostart services on active node, the standby node was able to start.
This should be perfectly safe for the running application (no downtime), but it is still advisable to do it off hours.
jsapello
6 Posts
0
November 17th, 2010 06:00
The IP is correct, and I have rebooted this node a few times - I am going to try restarting the agent on the active node tonight.
yito1
262 Posts
0
November 17th, 2010 06:00
Hi,
Please reboot this node or reboot AutoStart Agent Service and BackBone Service.
Is this address correct?
128.127.84.48
jsapello
6 Posts
0
November 17th, 2010 06:00
The IP is correct - I will give your suggestion a try - hopefully tonight after hours.
ecervant
63 Posts
1
November 17th, 2010 10:00
During your maintenance window, I would like to also suggest restarting the Backbone service and not just the Agent service on the good node. Since this appears to be a backbone communication symptom, it would be good to re-initialize both services on the good node. Once the services are restarted then start the backbone and agent service on the troubled node.
Before performing the restart on the good node, make sure the Agent and Backbone service are stopped on the troubled node. If the services are in a starting or a weird started state, this can cause problems when starting the services on the good node.