Veeam Plug-in for SQL - GFS retention

[dbeerts]

Hi,

I was wondering how you can become a sort of GFS retention with the SQL Plug-in. I was thinking about separate jobs for daily, weekly, monthly and yearly backups.
But then I read this on https://helpcenter.veeam.com/docs/backu ... ml?ver=120:
"Alternatively, if you enable retention policy for multiple types of backups of the same database, you must specify the same number of days to keep restore points in the backup chain for each backup type. Otherwise, Veeam Plug-in will keep restore points according to the lowest number."

As log backups are also needed, which will be kept for 2 days, what will be the behavior of the retention on full backups if configured as followed?
Daily: 14 days
Weekly: 4 weeks
Monthly: 1 year
Yearly: 7 years

[PetrM]

Hello,

There is no built-in GFS-retention for SQL plug-in but it's a valuable feature request for our future releases.

As mentioned in the user guide, there is no separate retention for full and log backups. The retention logic applies to the entire backup chain including full, differential and log backups. I would try the following workaround:
1. A daily SQL agent job for full backups and another one for periodic log backups: set 2 days retention in a daily full job. Thus, you will have log and full backups for 2 days
2. Create a separate image-level job with Copy-Only mode and GFS policy enabled

Thanks!

[dbeerts]

I was already afraid for this answer
GFS support for SQL plugin is a must have in my opinion. With the provided workaround you are still creating double backups and consuming twice the space required.
Any idea how long it would take to implement this option? I think it's relatively easy to do, as the plugin writes to separate .VAB files which could be flagged?

[PetrM]

Hello,

It's difficult to say now: sometimes features that seem to be easy to implement at first sight require a significant amount of engineering resources, and sometimes vice versa. I cannot comment on ETA but your request is noted.

Speaking about the workaround above, I recommend using the Capacity Tier to offload backups older than your operational restore window, thus saving space of the Performance Tier.

Thanks!

[Edward Parlakyan]

We are one of the largest Veeam customers in our country, and our organization has approved a policy for keeping long-term backups, which is an ideal fit for GFS settings. However, this data is SQL and SAP HANA databases, which we back up using Veeam plugins, since there are databases on the same server that are large but whose backups should not be included in the long-term policy, we need a more tailored solution.
I would be very grateful if Veeam added the ability to configure GFS settings in the SQL and SAP HANA Plug-in job settings. It would be a significant step forward for Veeam if the SQL plugin job could be created from the Veeam console itself, just like the SAP HANA plugin job, but with the GFS feature.

[david.domask]

HI Edward, welcome to the forums.

I've moved your post to an existing topic with the same request (for SQL plugin), but it is a known request. Thanks!

[Edward Parlakyan]

Hello David.

Thanks for the answer. I hope this feature will appear in the next versions.

[PetrM]

Hi Edward,

Yes, it is one of the most frequent requests for all application plug-ins but I cannot give any information about ETA. The main complexity for us is that we don't fully control the backup chain, as a plug-in we just provide an interface to stream data to Veeam infrastructure. Did I understand correctly that you'd like to see it as an additional setting of our application backup policy and not as a part of a standalone plug-in?

Thanks!

[dimaslan]

Hello all,

We have a customer on our datacenter for which we deployed two new DB servers on a High Availability Cluster, getting the databases storage from a third VM. After our research we found the only way to backup those is via the Veeam Plugin for SQL. We installed it and deployed a new VM and installed Veeam on it.
Backups worked during our testings, however this customer has a need for 30 daily, 12 monthly and 10 yearly backup retention.
Question is, where do I set up this retention and how? The plugin backups, they don't need then to go back more than a couple of weeks, and I am planning to run backups of my whole Veeam server. However, let's say they ask for a database from 6 or 24 months ago. How do I even get that?
I can see the plugin setting for retention is only in days. If I mount a backup of the Veeam server as it was 6 months ago, how will I be able to use that to get the backups restored?
Please help, we have to start backups for this customer right away.

Thanks.

[dimaslan]

Hello,

I have the same issue, only I am using the Veeam Plugin for SQL to backup a HA SQL Cluster with two nodes which are getting their SQL DBs storage from a third server.
The issue is this customer needs 30 daily, 12 monthly and 10 yearly backups and as it is I do not see how we'll be able to do that.
We have a backup repository on our Veeam server but it's on SSD so I cannot use that for LTR.
Can someone advise on how we can do this? I know we can setup separate jobs and the copy-only option does create a separate backup but I think the Plugin configuration can only point to a specific backup repository.

[PetrM]

Hi Dimitris,

I moved this post to this topic, it looks like your questions above are about the same functionality. GFS retention is one of the top requests but I cannot share ETA.

Technically, you can change the repository in the plug-in configuration wizard, run copy-only backups to another repository, then switch repositories back and run full backup to restart the backup chain. However, we did not test this scenario and I cannot guarantee stable functioning.

Would you consider the Capacity Tier as a temporary workaround? You can move backups falling out of the operational restore window to object storage.

Another idea is to carry out image-level backups with the Copy-Only mode of the third VM with DBs storage to enable GFS retention while the short-term retention can be managed by our plug-in. But I never tried this so I'd recommend testing all the possible recovery scenarios and making sure they seamlessly work in your environment before going this path.

Thanks!

[dimaslan]

Petr,

We specifically asked our customer to accept a sizeable quote for SSD storage since just one of their databases is 500 GB. This was causing backups for it to take a long time, restores too. All their databases are about 1TB.
This is most distressing that there is no option for us and I'm not sure what to do next. The Store server which provides the storage for the databases to the ha sql cluster can be backed up, however I don't think databases are visible there.
This is going to be a huge issue, we may have to move away from Veeam on this and just do native SQL backups to the SSD storage, which I think creates .bak files, then keep 10 years retention for those backups and be able to recover a .bak file when needed.
I am not sure even if I backed up my Veeam server, how I could bring back a backup chain from say a year ago.
Let's say I keep 30 days of full and differentials on the Veeam server (it's also a VM on our datacenter). If at some point we are asked to recover a database from a yearly of 2 years ago, how would we be able to restore it? I can see the Veeam plugin chains are in a proprietary .vab and .vasm file format. Can these once restored be imported somehow in Veeam and then be visible on the SQL server to recover from?