Data Backup and Recovery

SnapManager with NFS vmdk disks?

mvintinner
4,336 Views

I noticed that SnapDrive 6.4 now supports taking snapshots of VMDK disks that are on NFS volumes.  Does this mean that I can use my SnapManager products (Exchange & SQL) with a disk configured this way?

 

For example, in our current setup we have SQL running inside VMware vSphere 5.  The VM is running SnapDrive with the Microsoft iSCSI initiator from inside the VM to connect to LUNs.  Backups are performed with SnapManager for SQL.

 

What I'd like to do is create new NFS volumes for the SQL data, and attach VMDK disks from these new volumes to replace the iSCSI LUNs.  This would give me the benefits of RDM (in the sense that processing is handled by the ESX host instead of the VM) and the benefits of NFS (multiple IO channels, simplified datastore space management).

 

So question is:  will SnapManager for Exchange/SQL be able to snapshot this NFS volume as part of it's backup procedure?

1 ACCEPTED SOLUTION

radek_kubka
4,336 Views

Hi,

SMSQL could work with NFS VMDKs since SnapDrive 6.3.

(e.g. have a look at this: https://communities.netapp.com/message/41547#41547)

SME is still a no go on NFS, even with SDW 6.4 (and some people say it will never be due to Exchange requirements, but personally I think anything can still happen in a future )

Key thing about SDW 6.4 is, it supports Cluster Mode, so e.g. SMSQL can work for the very first time with Cluster Mode filers.

Regards,

Radek

View solution in original post

3 REPLIES 3

radek_kubka
4,337 Views

Hi,

SMSQL could work with NFS VMDKs since SnapDrive 6.3.

(e.g. have a look at this: https://communities.netapp.com/message/41547#41547)

SME is still a no go on NFS, even with SDW 6.4 (and some people say it will never be due to Exchange requirements, but personally I think anything can still happen in a future )

Key thing about SDW 6.4 is, it supports Cluster Mode, so e.g. SMSQL can work for the very first time with Cluster Mode filers.

Regards,

Radek

mvintinner
4,336 Views

Thank you.  This is exactly what I was looking for.

Danny_Makie
4,013 Views

mvintinner,

 

this is exactly what i was just researching. How did you go about placing the vmdk's on datastores then the flexvol.

 

Best Practice for iscsi is to have a 1 to 1 relation 1 Vol per LUn for each drive? 

 

Do i need 1 VMDK per Datastore per Volume? 

e.g SQL Database on VMdk sitting on it on Datastore and Volume. Separate Datastore and Volume for logs, 

 

or does is it not really necessart when using NFS.

 

Thanks

Public