Unsolved

This post is more than 5 years old

2 Posts

1046

April 8th, 2008 07:00

Backbone service will not start

Hi,

I'm using Autostart 5.2.1. Recently the second drive array on one of the nodes failed and had to be replaced. When the server was coming up it failed to start the backbone service. Node agent does not start as well but that is expected as it relies on the backbone. Transport and mirror services start OK.

I get the following error message when I try to start

ftbackbone.exe

Backbone panic on (pid 5856) Tue Apr 08 13:44:49 2008
in ISIS_MGT_PROTOS_MODULE pr_main.c/scan_sfile(), line 692
IP_FILE_OPEN: (62) could not open
ID00004608 sname sites: cannot open

I've compared the configuration of both nodes and found nothing. The heartbeat and mirroring links are up and running.

Any sugestions?

Regards,

Kamil

198 Posts

April 8th, 2008 11:00

Hi,


Have you tried demoting the server to an AutoStart secondary then promoting the server back to a primary ?


IKP

Message was edited by:
IKP

2 Posts

April 9th, 2008 02:00

Hi,

Demoting the server to secondary allowed me to start the backbone and agent, but when i promote it back to primary the agent and backbone service are restarted and do not come back.

Should I leave it at secondary, try to resync and then make it primary?

Regards,
Kamil

262 Posts

April 9th, 2008 03:00

Hi,

Generally, Agent Service and BackBone Service start even if Demote Node and Promote Node are operated.

It is likely not to operate correctly in your environment.

There are three methods.

1. Please look at the event log more. It is likely to become a clue.
2. It is necessary to reinstall AutoStart by the server that doesn't start BackBone.
3. It is necessary to request it to a technical support.

I'm sorry in a general answer.

Yoshinobu

198 Posts

April 9th, 2008 23:00

Kamil,


If this cluster has more than 2 nodes then yes you could leave this node as a secondary and promte one of the other secondries to a primary. If however, you only have two nodes then you will need to have both of them as primaries. If this is the case and you have tried all the obvious - verified connectivity across ALL the domain lines then there is a way with advise from support to delete the local database on this node and then restart - this will force the replicated database to be copied back to this node - if this fails then I would look to de-install EAS and then re-install.

Hope this helps

IKP

0 events found

No Events found!

Top