1 of 1 people found this helpful
Check your arerror.log file to see if arserver process lost connection to database for too much time and restarted.
When arserver restarts, it can take this kind of time (10 minutes), in the meantime you can't access the solution.
Anyway, check if arserver process stopped / started around the failover time, you should see it in arerror.log and armonitor.log.
How much time the SQL failover lasted?
Thanks Laurent. SQL failover only lasted a few seconds because it's clusted.
So, when failover occurs arserver takes 10 minutes to restart or can it reconnect instead?
Did arserver restarted?
Yes, and restart took 10 minutes.
I see, I'm no SQL DBA so well... Is there some kind of timeout on the SQL Server client you could play with?
How is configured the client?
SQL server had no issues because it's clustered so failover took a few seconds but Remedy had to restart
which took 10 minutes. So it has to be on Remedy.
I was rather speaking of the client actually
In the thread I found they tell that sometimes even if failover is successfull the client still tries to connect to the same server instead of the next.
I know they did something here for DB2, I'll ask them, should be kinda the same philosophy.
How do you prepare Remedy when database failover happen?
do you just let it restart which takes 10 minutes?
There is no special configuration for reconnection. What entries were posted in the arerror.log when this failover happened?
Mon Apr 09 15:29:53 390620 :Failure during SQL operation to the database (ARERR 552)
[DBNETLIB][connectionwtite send()). General network error ************
Mon Apr 09 15:30:33 390635 :Failure while trying to connect to the SQL database
Mon Apr 09 15:30:33 390635 :cannot initialize contact with SQL database (ARERR 551)
Mon Apr 09 15:30:33 Stop Server
Mon Apr 09 15:30:33 AR system server terminated -- fatal error encountered (ARNOTE 21)
Looks as if it didn't try to reconnect for long.
Ok, there was connection parameter so I'm good.
Out of curiosity, what is the parameter?
I seem to have a very similar issue.
We are on SQL 2008R2 SP1, Win2K8R2, Remedy 7.6.04 SP3
I have literally moved Remedy database from another SQL Cluster where the failover worked just fine (Similar environment as before but SQL wasn't patch to SP1 level).
Now the Remedy service is unable to re-connect after the failover, and it has to wait for the restart...
This is a last few lines from my arerror.log:
Sat Apr 28 17:32:24 2012 [DBNETLIB][ConnectionWrite (send()).]General network error. Check your network documentation. (SQL Server 11)
Sat Apr 28 17:32:26 2012 Dispatch : Timeout during data retrieval due to busy server -- retry the operation (RemedyServer) ARERR - 93
Sat Apr 28 17:32:41 2012 390620 : The SQL database operation failed. (ARERR 552)
Sat Apr 28 17:32:41 2012 [DBNETLIB][ConnectionWrite (send()).]General network error. Check your network documentation. (SQL Server 11)
Sat Apr 28 17:33:38 2012 390603 : Approaching physical stack limit. (ARERR 8749)
Sat Apr 28 17:33:38 2012 390603 : Unable to connect to the SQL database. (ARERR 551)
Sat Apr 28 17:33:38 2012 Stop server
Sat Apr 28 17:33:38 2012: AR System server terminated — fatal error occurred in ARSERVER (ARNOTE 21)
Sat Apr 28 17:33:39 2012 Dispatch : Cannot establish a network connection to the AR System server (RemedyServer (0) : RPC: Program not registered) ARERR - 90
Sat Apr 28 17:33:40 2012 : Action Request System(R) Server x64 Version 7.6.04 SP3 201201302357
(c) Copyright 1991-2011 BMC Software, Inc.
Sat Apr 28 17:34:40 2012 Dispatch : Cannot establish a network connection to the AR System server (RemedyServer (0) : RPC: Program not registered) ARERR - 90
Sat Apr 28 17:35:02 2012 AssignEng : AR System Application server terminated -- fatal error encountered (ARAPPNOTE 4501)