יום חמישי 19 אפריל 2012 08:40
I need some advise on how to best design the following.
We need a firmware upgrade of some internal datacenter switches. What happens is that the Default GW and the connector to our Geo Clusters will be disconnected for like 30 secs.
Its 2 Nodes Windows 2003 Clusters most of it but also Win2k8R2.
What i need is to move all Cluster groups to 1 node and then put this node in like a service state where it's allowed to loose the quorom.
So 1 node is up. 1 node will be paused and the FSW will be gone for like 30 secs.
How do i manage this the best way.
Whould it be possible to tweak the MNSFileShareCheckInterval as per http://support.microsoft.com/kb/921181
- נערך על-ידי RFalken יום חמישי 19 אפריל 2012 09:56
שבת 21 אפריל 2012 07:36מנחה דיון
I like to play thing save and if you can I turndown the cluster nodes it is just like a reboot. Bing one node up and see if things are running fine , and bring the othe node up. If you can't do this and the cluster needs to be up all the time. then you can pause the nodes. but if you need more time for the switches and rollback the firmware or other things. then a save shutdown is the thing I should do.
If you have troubles with the clusters just because you tweaked some settings you will be sorry afterwards.
Greetings, Robert Smit [MVP] http://robertsmit.wordpress.com/ “Please click "Vote As Helpful" if it is helpful for you and Proposed As Answer”
יום ראשון 22 אפריל 2012 15:15בעלים
As Robert said, all you need to do here is Pause the nodes, drain them, then patch. No need to modify the cluster properties you list above. See this KB article, it outlines the process:
We have a new Node Maintenance Mode feature in Windows Server 2012 that will automate and simplify the process. See this blog: http://blogs.msdn.com/b/clustering/archive/2012/04/03/10290554.aspx
יום שלישי 24 אפריל 2012 13:55
For best particles before the maintenance activity, Move the cluster resources and check all the cluster resources are working in both the servers.
IF everything is works fine in both nodes, manually move all the Cluster resource to another node before started upgrading firmware of some internal datacenter switches.
Regards, Sridharan. M
- נערך על-ידי Sree8133 יום שלישי 24 אפריל 2012 14:04