Re: Active/Passive Node Cluster DR Resources

Tech-Archive recommends: Speed Up your PC by fixing your registry



Would like to consider the following scenarios:
Loss of single node & restore from backup to same hardware

With a loss of a single node, the other node has the data and resources, in
this case we'd not bother with a restore, instead we'd evict the failed
node, repair the failure, hardware or whatever and if necessary reload the
OS. Once we had the failed Node working we'd rejoin the cluster.

Loss of single node & restore to different hardware

Same as above, we'd evict the node, and then join the new hardware to the
cluster. Before we did that however we'd ensure that the Shared storage was
presented to the new node.

Complete loss of cluster (both nodes) - rebuild and restore data from
backup

Our priority would be to get either node back, and connected to the shared
storage and the data available. We wouldn't necessarily be interested in
whether the node was still clustered or not at this point. Once we had the
data back to the users we'd fix the 2nd node, and schedule a maintenance
window to connect the shared media to the cluster. Note that the Quorum
drive wouldn't necessarily need to be connected to the 1st node, which would
enable us to get the 2nd node clustered and working.

We're looking at HP Storage Mirroring to handle failover to a stand alone
machine at a remote site. Virtual machines are being considered as well.

Loss of shared storage device

Repair or replace the shared storage device and recreate the LUNS and
present to the Cluster. Then start the restores of the data to the shared
storage. This would be ugly for us, we're currently looking at HP Storage
Mirroring to help in this event, where we'd fail over to a DR Machine at a
remote site that would have a Mirror of the clusters data. Assessing band
width requirements at this time, and whether failover should be manual or
automatic. Also looking at the use of Virtual Servers at this site instead
of Physical machines.


.



Relevant Pages

  • Re: Error message question: The physical disk cannot...
    ... for the internal disks and one for the shared disks. ... >I received the following message today when configuring Cluster ... > storage bus as the boot disk. ... > 1) Is this a hardware issue that can only be resolved by new hardware ...
    (microsoft.public.windows.server.clustering)
  • Help-cluster restore on Windows 2000
    ... we follow below procedure to restore our cluster if storage is ... the cluster sometimes didn't regconize the hard disk ...
    (microsoft.public.windows.server.clustering)
  • Re: Disactory recovery from MS SQL clustering
    ... Scenario 6—Single Cluster Node Corruption or Failure ... If the Event Log indicates that the cluster database on the local node is merely corrupted, you can perform a System State restore on that node to replace the local cluster database. ... After verifying that all cluster resource groups have been successfully moved to other nodes, repair or replace the failed hardware. ... see To restore files from a file or a tape and Scenario 8 below. ...
    (microsoft.public.sqlserver.clustering)
  • RE: Help-cluster restore on Windows 2000
    ... Actually we assumed the two cluster nodes and storage ... recover from backup instead manually cluster re-creation. ... >> that after restore, the cluster sometimes didn't regconize the hard disk ... >> Best Regards, ...
    (microsoft.public.windows.server.clustering)
  • RE: Server 2008 Cluster Issues
    ... I also upgraded the storage MPIO drivers on the Windows 2008 server to 10.36 ... We are planning to make the same upgrade in couple of days in the production ... Cluster resource 'FileServer-' ... the Cluster administrator will hang and Windows ...
    (microsoft.public.windows.server.clustering)