Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Drive Locked when testing fail over

2 views
Skip to first unread message

Amador

unread,
Feb 9, 2010, 9:56:31 AM2/9/10
to
Everyone,

I have configured a SQL 2005 cluster with two nodes. My SQL data is on a SAN
attached drive managed by NetApp which is drive x:. When I try to test the
fail over or move from one node to the other x: drive is locked by the first
node and the fail over fails. Can anyone share any pointers why this is
happening.

Thanks

John Toner [MVP]

unread,
Feb 9, 2010, 10:06:32 AM2/9/10
to
Can you explain what you mean by "locked"? Does the disk ever go offline?
Just wondering if its failing to go offline, or failing to come online on
the other node.

On the node where the disk is failing, you'll want to look at the
c:\windows\cluster\cluster.log (assuming Win 2003) file to see why the disk
is failing to go offline/online. This log is based in UTC time zone, so
you'll need to adjust accordingly when looking through this log.

Regards,
John


"Amador" <ama...@agconsulting.info> wrote in message
news:%23ldgNgZ...@TK2MSFTNGP02.phx.gbl...

Amador

unread,
Feb 9, 2010, 3:26:21 PM2/9/10
to
basically is failing to offline and switching over to the other node. At
that point the cluster is useless and I have to remote into the server and
start it again. Looking at the cluster.log from yesterday and today I don't
see much.

"John Toner [MVP]" <jto...@DIE.SPAM.DIE.mvps.org> wrote in message
news:uapv4lZq...@TK2MSFTNGP06.phx.gbl...

John Toner [MVP]

unread,
Feb 12, 2010, 9:46:21 AM2/12/10
to
Are the other resources going offline successfully? Are there any 1069
errors logged in the System event log?

"Amador" <ama...@agconsulting.info> wrote in message

news:OijshYcq...@TK2MSFTNGP06.phx.gbl...

Amador

unread,
Feb 12, 2010, 1:13:36 PM2/12/10
to
Yes many 1069.

"John Toner [MVP]" <jto...@DIE.SPAM.DIE.mvps.org> wrote in message

news:euAmlI$qKHA...@TK2MSFTNGP06.phx.gbl...

Amador

unread,
Feb 17, 2010, 11:09:26 AM2/17/10
to
John,

Any Ideas on what I should be looking at to get this resolved?

John Toner [MVP]

unread,
Feb 17, 2010, 4:05:26 PM2/17/10
to
Which resources are reporting the 1069 errors?

What are the events in the cluster.log at the time of these 1069 errors?


"Amador" <ama...@agconsulting.info> wrote in message

news:O3f%23Uu%23rKH...@TK2MSFTNGP05.phx.gbl...

Amador

unread,
Feb 18, 2010, 11:02:39 AM2/18/10
to

So basically this is what I notice. I have a resource type called SQL Server
(Altiris) and when looking at the under possible owners I only see one node
and not the second node called NY-SQL2, I only see NY-SQL2. All other types
have both nodes as owners.

"John Toner [MVP]" <jto...@DIE.SPAM.DIE.mvps.org> wrote in message

news:exwNvTBs...@TK2MSFTNGP05.phx.gbl...

John Toner [MVP]

unread,
Feb 18, 2010, 2:42:18 PM2/18/10
to
That will certainly prevent the group containing that resource from failing
over to your NY-SQL2 node.

"Amador" <ama...@agconsulting.info> wrote in message

news:eD8SKPLs...@TK2MSFTNGP06.phx.gbl...

Amador

unread,
Feb 22, 2010, 1:10:56 PM2/22/10
to
John,

Any ideas how I could fix this without having to rebuild the cluster. I will
like to get this resolved but not sure how to tackle it.

"John Toner [MVP]" <jto...@DIE.SPAM.DIE.mvps.org> wrote in message

news:OWyV8JNs...@TK2MSFTNGP02.phx.gbl...

Geoff N. Hiten

unread,
Feb 22, 2010, 3:35:06 PM2/22/10
to
It sounds like the SQL instance is not installed to the NY-SQL2 node. Check
for local binaries on that node.

If so, add the node to the possible owners list and try a "move group". Be
prepared to fail it back if necessary.

--
Geoff N. Hiten
Principal SQL Infrastructure Consultant
Microsoft SQL Server MVP


"Amador" <ama...@agconsulting.info> wrote in message

news:O0K5dp%23sKH...@TK2MSFTNGP05.phx.gbl...

Amador

unread,
Feb 22, 2010, 4:55:15 PM2/22/10
to
So install the default instance on the server, I just looked in add/remove
and clicked on MS SQL server 2005 and there is nothing under SQL server 2005
Instance. There is no instance in the secondary node.

"Geoff N. Hiten" <SQLCra...@gmail.com> wrote in message
news:uD8BS6$sKHA...@TK2MSFTNGP05.phx.gbl...

Geoff N. Hiten

unread,
Feb 22, 2010, 10:03:56 PM2/22/10
to
I think we have found the problem.

--
Geoff N. Hiten
Principal SQL Infrastructure Consultant
Microsoft SQL Server MVP

"Amador" <ama...@agconsulting.info> wrote in message

news:urJU0mAt...@TK2MSFTNGP05.phx.gbl...

Amador

unread,
Feb 23, 2010, 10:36:25 AM2/23/10
to
so just reinstall sql on the second node and make sure the instance is
installed?

"Geoff N. Hiten" <SQLCra...@gmail.com> wrote in message

news:elL5jTDt...@TK2MSFTNGP05.phx.gbl...

Geoff N. Hiten

unread,
Feb 23, 2010, 11:16:28 AM2/23/10
to

Install is cluster-wide so you will have to re-run the installer on the
first node and have it add the second node to the SQL instance.

See "Maintaining a Failover Cluster" topic in BOL.

--
Geoff N. Hiten
Principal SQL Infrastructure Consultant
Microsoft SQL Server MVP

"Amador" <ama...@agconsulting.info> wrote in message

news:%23MdMy3J...@TK2MSFTNGP05.phx.gbl...

0 new messages