SQL Backups

[kuhn]

I am curious to know how everyone is backing up SQL. We were once told using an agent was a better choice for heavily transactional databases to eliminate a snapshot "stun". So, we setup agent level backups and transaction log backups occurring every 15 minutes. Is this still a thing or would using application aware backups be a better way of do this? Trying to weigh the options to see what others are doing.

[jason.berry]

The answer to this question is more preference /simplicity if your SQL servers are sensitive to the snapshot stun of your hypervisor or you need to support clustering for SQL Server then agent is a great option. Either backup methods use application aware processing in order to do log truncation etc.

However, if these are systems that are managed by DBAs and they want to the ability to control how the databases and logs are backed up and have their own scripts using the native SQL Server tools to complete those backups, then you might consider Enterprise Plugins as it has even more flexibility and scalability as it can take advantage of parallel streams that the agent/hypervisor can't and can be leveraged in the DBA's existing scripts.
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

[PetrM]

Hello,

To elaborate a bit more on it, you can find additional information in this FAQ. Also, please keep in mind that VSS does not support volumes larger than 64 TB, in this case the plug-in will be the only option.

Thanks!

[geetansh]

Hello,

One question on the sql plugin backups, can we configure it to automatically exclude the offline databases?

[PetrM]

Hello,

No, there is no such option at the moment. However, you can explicitly exclude specific databases in the plug-in backup wizard.

Thanks!

[geetansh]

Thank you. I have submitted a feature request for it in a separate post. It will be super helpful for the DBAs.

[PospiKM]

Hello geetansh,

please could you share link with the submitted requested?

From my own perspective, I would warmly welcome any possibility how to automatically exclude the offline databases from the job. Currently it's a big problem for us, having more than 200 SQL servers with randomly offline databases which are totally out of our controll due too many reasons and of course causing the "job failure". And we cannot exclude it manually in the jobs as it does not make any sense for us.

Thank you.