File Share Witness doesn't work (Full Version)

All Forums >> [Microsoft Exchange 2007] >> High Availability



Message


SilverICE -> File Share Witness doesn't work (15.Feb.2007 5:21:04 PM)

I've done something wrong setting up my CCR and my File Share Witness isn't working.  When I reboot or shut down one of my cluster nodes the entire cluster goes down until the other node comes back up.  So it seems I don't have a majority when one of the nodes is down. 

From a command prompt on both nodes I can successfully run 'cluster res "Majority Node Set" /priv'.  They both point to the correct path of the share on my HT server.

Any clues on how to troubleshoot this?




Henrik Walther -> RE: File Share Witness doesn't work (16.Feb.2007 2:42:33 AM)

And you configured the permissions on the File Share Witness share correctly?

Has any content been create in the respective share?




SilverICE -> RE: File Share Witness doesn't work (19.Feb.2007 4:44:57 PM)

Yes, I've double checked the permissions per your CCR build documentation, and have even opened it wide up to everyone.  I can browse to the share from the exchange server and there is content in the foler.  Any ideas before I wipe and start over?




Henrik Walther -> RE: File Share Witness doesn't work (20.Feb.2007 4:08:52 PM)

Hmm nope not really [8|]




SilverICE -> RE: File Share Witness doesn't work (21.Feb.2007 10:33:05 AM)

Just for my own clarification: a CCR cluster should fail over just like a single copy clustered Exchange 2003 server with shared storage, correct?  To test failover, I should be able to restart the active node which would initiate failover, right?




Skywalker -> RE: File Share Witness doesn't work (20.Dec.2007 5:18:36 AM)

quote:

ORIGINAL: SilverICE

I've done something wrong setting up my CCR and my File Share Witness isn't working.  When I reboot or shut down one of my cluster nodes the entire cluster goes down until the other node comes back up.  So it seems I don't have a majority when one of the nodes is down. 

From a command prompt on both nodes I can successfully run 'cluster res "Majority Node Set" /priv'.  They both point to the correct path of the share on my HT server.

Any clues on how to troubleshoot this?


I have the exact same problem.

I'm sure that my File Share Witness works. I can clean the directory en when I do a failover it fills up with the expected files again. Rights are ok.
'cluster res "Majority Node Set" /priv' show me the correct path. I'm using a cname record and i've set the "DisableStrictNameChecking" registry key.
If i do a manual failover for the exchange group and the cluster group all works fine.
But when I reboot the HUB server wich has the File Share Witness on it, my cluster crashes. When the HUB server is back online everything comes back online. Same thing happens when I pull the public network cable from the active node. The cluster service on both cluster nodes is stopped until I plugin the network cable again. Then it initiates the failover the the other node.

I'm going crazy finding a solution for this.




Page: [1]