IBM Storwize V7000 - MetroMirror vs. MetroCluster

[petter_glenstrup]

Hi

Does any of you have any information on what the advantage of MetroCluster is over IBMs MetroMirror ?

As i can read the product only use FC, and they do not have Dedupe and the Snapshots are CoW.

But when a potential NetApp MetroCluster, that will only use FC, will this product do the same as we can do with MC ??

[yarosh]

V7000 MetroMirror is more analogous to SnapMirror. The value in MetroCluster is the automated failover and failback without anyone in the middle.  It's also deisgned to prevent the split-brain scenrios that come up when software is used to automate the automatic failover.

[paulstringfellow]

Can only support what has been said, MetroCluster is a pretty clever NetApp function, allowing for no single point of failure design, in a stretched data centre environment, allowing for the controllers to be separated over great distance and presenting a single storage controller.

IBM don’t have this functionality yet within V7000, although V7000 is a nice solution especially its storage virtualisation capability, built on the IBM SVC unit, however in terms of a smart storage solution, it offers nowhere near the capability you will get with a NetApp filer as you have already mentioned, no dedupe, copy on write snapshots, however they will support 255 snapshots…but it does strike me as a classic example of wanting to be a NetApp solution.

But quick answer to your question, if you want a metrocluster solution, then V7000 not the answer and…well metrocluster is! ☺

[radek_kubka]

Yeah, I won't bring anything new to the table, yet (funnily enough) both technology names are 100% spot on in this case :

- MetroCluster = a cluster stretched over a metropolitan distance

- MetroMirror = mirroring (synchronous) over a metropolitan distance

Regards,
Radek