Windows 2008R2 Fail Over cluster, what if the shared disk fails?

[lks]

Hi guys,

I had configured failover clusters on some testing scenarios. 2 nodes and one machine acting as a shared disk using MS iSCSI software target, all are virtual machines under hyper-v. It works well, But what it the shared disk fails? how can I make it redundant?

For example, let's assume this scenario:

*the shared disk may be a server with iSCSI software target, or a NAS device...

No matter how many nodes, if the shared disk fails it will make everything to fail, bad thing. :biggrin:

But, if I have this:

how can I make both disks redundant? I mean, if one shared disk fails, is it possible to make the second shared disk to take place automaticaly and with the same data?

Am I thinking this wrong and I have to plan it different to reach a high availability?

 

[imagoon]

Shared disks are typically on replicated SAN's for HA environments. Sometimes they span datacenters as well.

Example would be buying 2 NetApp 2040's, configuring replication so all disk writes on one is replicated to an "offline spare." Should the master fails, the spare pops online and takes over the requests.

This sync can be done "periodically" meaning say "1 hour of work is lost" or "fully synced", where the SANs don't report the disk write as successful until both SANs have confirmed the write as committed.

 

[dphantom]

Shared disks are typically on replicated SAN's for HA environments. Sometimes they span datacenters as well.

Example would be buying 2 NetApp 2040's, configuring replication so all disk writes on one is replicated to an "offline spare." Should the master fails, the spare pops online and takes over the requests.

This sync can be done "periodically" meaning say "1 hour of work is lost" or "fully synced", where the SANs don't report the disk write as successful until both SANs have confirmed the write as committed.

The latter is what we do with our EMC SAN. 2 data centers mirrored, fully synced for our critical data.