/* Font Definitions */
panose-1:2 4 5 3 5 4 6 3 2 4;
mso-font-signature:0 0 0 0 0 0;}
panose-1:2 15 5 2 2 2 4 3 2 4;
mso-font-signature:-1610611985 1073750139 0 0 159 0;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
font-family:"Times New Roman","serif";
margin:1.0in 1.0in 1.0in 1.0in;
I have some questions in regards to features of the MD3000i.
The Disk Copy and Snapshot
feature with this SAN. I wanted some more clarity on how to use these features
in a DR scenario. If I mirror, snapshot or Disk Copy a Disk Group during a
failure the SAN will not point to the other source, correct? Example, If I
mirror two drives and the active drive fails, how would I get my host to point
to the other drive?
Also, with Snaps and Virtual Disk Copies they only perform
those functions within the Disk Group so what can I do if the Disk group fails?
The san takes care of drive redundancy. If a mirrored pair has a drive failure the current host continues to run, even during the rebuild once the failure is replaced. Disk Copies if used for redundancy should be made to a separate disk group / set of drives. You can boot up and run on a snapshot, but the snapshot always needs the original VD it was made from up and running
Generally with a hot swap spare defined, the disk group / mirrored pair / raid 5 / raid 6 / raid 10 stays up just fine duyring the rebuild.
Now for disaster recovery or even just to achieve some enclosure protection? This one is tough on the MD3000i with no SAN to SAN copy / replication. What we do, is take a snapshot, then disk copy our disk group(s) using the snapshot as the source to an MD1000 expansion chasis. The disk copy is 100% run-able / bootable, no restore needed back to the date / time you took the snapshot.
We use Evault backup to replicate our data to our DR location. Evault puts the data back on the MD3000i housed at that location, avoiding the no san to san limitations.
Many ways to go with this, using ESX as well, but this is how we do it.
More expensive sans have san to san copy / replication and it is wondefull, but you pay like 6 times as much for it over the MD3000i.
How would you restore a virtual disk from a snapshot virtual disk in the event of lost data?
When you say "lost data" are you talking data corruption, where the virtual disk (diskgroup and virtual disk) is still ok, or are we talking a double drive failure (or single drive failure on raid 0)?
Member since 2003
Snapshots rely on access to the source virtual disk as only the changed blocks are coped to the snapshot repository, as such it all depends on what data was lost. If data was overwritten then the original could be obtain from the snapshot.
The typical usage scenario is to use the snapshot as the source of a backup operation so that the original source virtual disk (i.e. the one that the snapshot was taken of) can remain online and fully available.
Can you provide a specific example?
Suppose you have a snapshot of a source virtual disk and a folder is deleted from the source virtual disk, what is the best way to recoved the folder from the snapshot repository?
I'm talking about corrupt or accidental deletion of data from the source virtual disk where the disks are OK.
In those cases the snapshot virtual disk can be mounted and the original data at the point in time that the snapshot was taken can be read.
Please understand that snapshots do utilize additional capacity and that the repository space has to be managed.
That is what I thought. Thanks Dave.