looking for use snapcenter to replace tape backups for Exchange, MS SQL and VMware

[FelixZhou]

 We are setting up NetApp AFF300 in our data center and DR site. I heard we can use snapcenter and snapvault to eliminate the backup of NDMP for cifs shares use snapshot and snapvault combination  replece backup solution for Exchange, MS SQL and VMWare. Can anyone please share your experience. thanks in advance.

We are backing up a few hundred TB data from Exchange, MS SQL, VMWare and CIFS shares. their backups and  restores cost us a lot of time, effects and money, if snapshot for short term and snapvault for long term can become our data protection solutions. then NetApp will make a huge change.

is that really true we can restore Exchange emails/inboxes, SQL databases and VMWare VMs from snapshots? on EMC VNX, data recover from snapshots did help us a lot.

looking forward to hearing from you....

[Ontapforrum]

Hi,

SnapCenter does a very good job of backing SQL/Exchange/VMware stuff, we are currently backing up Oracle, SQL & VMware. Only one left is 'Exchange' which we will move end of this year. CIFS is simply vaulted to secondary storage and we keep it for longer duration.

Vaulting (Secondary Storage) is for longer extended retention is a great tool and has been there for ages.
CIFS/NFS NDMP is totally independent of SnapCenter. Standard replication stuff filer to filer, if you can keep them on disk (Vault destination filer) for longer duration then you don't need tapes. You can even tier them to CLOUD.

Features of SnapCenter Plug-in (This is covered in the documentation resources):
• User can take consistency group snapshots of storage volumes across ONTAP clusters.
• Users can leverage pre and post Scripting framework to backup custom applications
• Users can update snapshots taken on the primary to an ONTAP secondary leveraging the existing replication relationship (SnapVault/SnapMirror/Unified Replication), using a SnapCenter Policy. The ONTAP primary and secondary can be ONTAP FAS/AFF/Select or ONTAP Cloud.

SnapCenter document resources:
https://www.netapp.com/us/documentation/snapcenter-software.aspx

For more info, you can reach out to forum or the support.

Thanks!

[FelixZhou]

good information.

A few more questions you may help. thanks

1). for SQL, VMware and Oracle, usually how long you keep the snapshots for short term protection locally ? How many percentage will the snapshots  consume on volumes? and will you recover VM and database directly from snapshot to overwrite the existing oned or make a new copies?

2). for snapvault long term protection, usually how long will you keep the data, weeks, months or years? 

3). for long term snapvault, if the destination cluster is on remote site, will it cause too long to recover a big amount of data?

[Ontapforrum · Accepted solution]

In general, 'Data protection' is nothing but taking data located in one location and making a copy of it in a different location to serve two use cases:

• Backup : The objective is to restore from the secondary to the primary with no intention of failing over
to the secondary. This implies that the primary purpose of the secondary is archival storage. Therefore, you might have more data in the secondary than in the primary.

At our place:

In Primary: Irrespective of Protocols/application, we keep 30 days worth of data
In Vault : 3 Months, and for exchange it is beyond (I think it is governed by standard data retention required by the Exchange compliance policy).

• Disaster recovery (DR) : An exact replica or copy is maintained in the secondary and used for fail-over from the primary to the secondary if there is failure at the primary site. This is actual DR copy : Mirror Image of the Primary file-system. If DR systems must be brought online, SnapMirror relationship can be broken, which makes the destination volumes read/write and ready to use. SnapMirror allows you to synchronise the original source with the changes made at the destination and then re-establish the original SnapMirror relationship.

This is the latest TR (Mar,2020) that covers some useful information:
https://www.netapp.com/us/media/tr-4015.pdf

Wrt this query : for long term snapvault, if the destination cluster is on remote site, will it cause too long to recover a big amount of data? Well, it all depends upon your WAN connection/speed (as far as pipe is concerned), especially with network compression etc it is not too bad. To be honest, it been used by  thousands of customers and it is a solid dependable technology. However,  testing is very important to simulate such events.  Half yearly testing is not  aggressive and achievable with proper planing . Keep reading and keep looking, you may find lot of  useful information/experiences on netapp as well as other netapp forums where customers have shared their views.

Thanks!