sql and veeam query

[terrance101]

Hey guys,

got another query here. When it comes to my veeam replications I am not taking the backed up sql databases across but we're going to be doing a replication every 8 hours and copy the logs across to the remote site every 15 minutes. It looks like you have to have the last backup in sql to then apply logs post last backup to get it back to most current state. The problem is we're looking at like 30 gigs for just one set of the last backups. Has anyone found encountered this issue with veeam and sql and have you found any trick to getting around having to copy so much data across to the remote site?

[veremin]

I’m wondering whether 30 GB in question applies to SQL logs or to the amount of changes that needs to be copied by VB&R during each incremental pass.

Thanks.

[terrance101]

the 30GB is the final backup of each of the sql databases. The logs come to around 1GB a day which isn't too bad. I have excluded the backup directory of the sql backups during replication because that would bring the 30GB across everyday. Basically there must be a trick with sql where you don't need the final backups and therefore the extra 30GB is anyone has come across this before.

[Vitaliy S.]

Hello,

Exchange and SQL Server VMs will always generate large increments because these applications are highly transnational. This means that a lot of file system changes in the VM virtual disk will be generated, so there is no "special" magic/steps that should be taken to optimize the size of the incremental changes for these VMs. However, what you can still try to do is to re-schedule/re-configure some maintenance tasks (if any) in SQL Server, should reduce the backup size.

Btw, I didn't quite get if you're using replication jobs or backup jobs to protect your SQL server. Can you please clarify?

Thank you!

[emachabert]

Vitaly, I think Terrance is speaking about SQLServer based backups (scheduled maintenance plan running a backup task to file), which is, as I understand, creating a 30 gig .bak file.

Terrance why not replicating more often with Veeam ? the longer you wait between replicas, the more you have to copy to the other site.
If you need to reduce your RPO to 15 mins, perhaps you'll need to switch to a log shipping based SQL cluster ?

[Vitaliy S.]

Hello Eric,

Yes, could be In this case another alternative would be to place these bak files on the separate drive and then exclude it from the backup/replication job.