June 2, 2009 at 2:30 am
Can anyone provide a link or instructions on how to install SQL 2008 SP1 on a 2node active/passive cluster.
do i just install on one node then the other or are there steps to take first
thanks
June 2, 2009 at 7:03 am
This is one of the awesome things that Microsoft has done with 2008, and really something that has been missing up until now.
The first thing is to download the service pack (by the way do this for your Cumulative Updates also).
Once downloaded apply the service pack to the Passive Node (yes, the passive one). Perform a reboot.
Failover the cluster.
Tada SP1 is installed and now in use.
Go ahead and deploy on the other node, as you are in active/passive you don't even need to fail back.
Down time? Yeah, under 5 minutes.
Of course, as always, performs backups prior to deploying the Service Pack, and if you have a test cluster run through it on there.
After the installation of SP1 I would recommend installing SP1 CU#2 and testing that in your development/testing area.
June 2, 2009 at 2:57 pm
To install SQL 2008 it's not needed to install it on the both nodes. Why the SP1 has to be installed on the both nodes. I found a Microsoft document about this installation but did not yet put it in place. There is no downtime. Here it is: http://support.microsoft.com/kb/958734. Everyhting is done on the passive node if I well understood.
What do you think about it ?
June 2, 2009 at 9:08 pm
Not applying SP1 on both nodes enables you the ability to go back should you run into a problem. It also minimizes the downtime for patching. Should you patch both machines at the same time then you would have to put SQL server offline to do do.
Downtime is as long as it take SQL to failover in the cluster and bring the databases back up. Even that process is downtime, especially as you will want to check the SQL logs before allowing users to connect to the instance.
June 3, 2009 at 2:32 am
Thanks for your replies, it is now installed and working fine.
thanks
November 26, 2009 at 9:38 pm
To apply Service Pack 1 on SQL Server 2008(Outage will be minimal):
1)Failover Group DBQA to DBSQ02 Node
2)Apply SP1 on DBSQ01 Server(Including Database Engine and Analysis Service)(Passive Node)
3)Reboot the DBSQ01 Server
4)Failover All group to DBSQ01 Server
5)Apply SP1 on DBSQ02 Server.(Passive Node)
6)Reboot the server DBSQ02.
7)Finally Move ASQA_PR and ASQA_RPT to DBSQ02.
Will above plan successfully apply the SP1 on SQL Server 2008?
August 23, 2010 at 5:36 pm
Do we need to stop the SQL Server services before applying SP1 in Cluster environment? Because , we have only Single node.
August 24, 2010 at 2:45 am
SQL Server Service should be in running mode. But there should not be any connection to the server for either of node.
August 24, 2010 at 4:57 am
pshaship (8/23/2010)
Do we need to stop the SQL Server services before applying SP1 in Cluster environment? Because , we have only Single node.
Not normally, no! But a cluster with only one node will force a reboot when you apply the service pack.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
November 24, 2011 at 6:57 pm
SQL Server Service should be in running mode. But there should not be any connection to the server for either of node.
Is this correct? I thought on active node and can accept connections and function normally while we install on passive. Can someone clarify
Regards,
Raj
November 25, 2011 at 12:58 am
arr.nagaraj (11/24/2011)
I thought on active node and can accept connections and function normally while we install on passive. Can someone clarify
That is correct.
The forums are full of good and bad advice, you need to look to more experienced users posts for clearer answers.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
August 20, 2012 at 9:55 am
Do a rolling upgrade.
1. Isolate primary node from passive so it can't failover
2. Apply service pack to secondary/passive node
3. Activate passive node back on cluster for failover
4. manually failover primary to secondary
-- at this point you have upgraded one node
5. Isolate secondary (now active upgraded node) from cluster
6. Apply service pack to primary inactive node
7. activate the node back in the cluster
8. manually failback to the primary
9. Take system db backups
Takes about 5 minutes each node.
No reboot of host required.
SQL server is online on node you are not upgrading. Failver applies the upgrade.
Viewing 12 posts - 1 through 11 (of 11 total)
You must be logged in to reply to this topic. Login to reply