ONTAP Discussions

DR procedure for Snapmirror with (CIFS, NFS,iSCSi,FCP)

vreddypalli
4,806 Views

Hi,

Is there any document from NetApp which clearly defines the steps to be followed when a disaster occurs on the source for all the protocols..?

Please share any URLs or documents.

Thanks in advance.

Regards,

Vijay

5 REPLIES 5

radek_kubka
4,806 Views

Hi Vijay,

IMHO, DR failover is always host & application-specific (with the exception of file shares, where hosts aren't involved).

For CIFS, you can utilise MultiStore DR:

http://now.netapp.com/NOW/knowledge/docs/ontap/rel80/html/ontap/vfiler/GUID-B8F3DFCE-D749-472C-BFB7-0D140B3857EA.html

For VMware, you normally would be looking at SRM:

http://www.netapp.com/templates/mediaView?m=tr-3822.pdf&cc=us&wid=87646022&mid=32231052

Regards,

Radek

bertaut
4,806 Views

Hello Vijay,

Your question is not very specific. Follow the link below for a list of guides worthy reading to address your question. They are for the ONTAP 8.0.2 release;

http://now.netapp.com/NOW/knowledge/docs/ontap/rel802/

  • Data Protection Online Backup and Recovery Guide
  • File Access and Protocols Management Guide
  • System Administration Guide
  • Storage Management Guide

A basic example of process to follow in the event of disaster:

  1. Local disaster affecting only one node in a cluster - Takeover of failed node in the cluster takes place, all the CIFS/NFS/ISCSI/FCP data are managed by the other node in the cluster in a nearly seamless fashion (CIFS/NFS data might experience a slight bleep during takeover). Once you address the failure on failed node, initiate giveback.
  2. Local disaster affecting both nodes at primary site - Snapmirror should be used in this instance. Break snapmirror relationship, have end-users point CIFS/NFS data to secondary storage devices. For the iSCSI/FCP data, you may need to quiesce application before mounting snapmirrored data or rely on software such as SnapManager.

Going over the guides listed above would be a great start.

Best of luck.

vreddypalli
4,806 Views

Hi,

Thank you very much for the respone.

We have to submit a DR procedure document to the client running CIFS.  We know the general steps, but customer is looking for some document from netapp.

I have already referred the above documents, there is no example given for a DR setup.

The customer site doesn't have multistore license(no vfilers).

Regards,

Vijay

aborzenkov
4,806 Views

DR involves really far more than just storage failover. So it would be misleading for a storage vendor to claim it offers complete DR setup procedure. Such document belongs to a project and is usually done by company that implements project (or by customer itself).

We had several DR projects and each one was different and accompanied by custom tailored documentation.

bertaut
4,806 Views

It goes without saying that any DR scenario involves far more than just the storage failover; the guides provided were as I stated in my previous post a great start. Any complete DR would require careful planning and synchronization from the Storage, Networking, Application, VM, end-users teams etc... The DR SOE will differ from one environment to another - the example provided above is merely an example of steps performed from the storage perspective in the event of DR.

Regards.

Public