SQL log to large to fit within log period

[mrholm]

Hi
We have Veeam 11 and see some behaviour that we want to ask if there's any way to come around.
All SQL databases in Full Recovery are backuped with a period of 60 min but sometimes some of the logs are to big to be able to be backuped within the 60 minutes and then Veeam stops the current log backup and start a new interval.
Is it possible to instruct Veeam to contine the current backup even if it takes more than 60 minutes so we get a log backup insted a lot of broken ones?

[HannesK]

Hello,
if the log changes are so large, that they can't be copied, then that sounds like a performance issue. Continuing to copy would just make the problem larger as far as I see.

What is the reason from your point of view, that it takes longer than 60min to copy? CPU (for the compression of the logs)? Network? IOPS?

Best regards,
Hannes

[mrholm]

No it's not performance issue, we have large development systems and they do regular restores of databases and then a lot of "read in" to the databases creating large log files. Then we are not able to perform the log backup and the log file then continue to increase during the day.

[HannesK]

still sounds like a performance issue to me.

if there is one "spike" where much data is generated and the rest of the day there is now change rate, then increasing the log shipping window comes into my mind. if that's not an option, then more performance is needed to stay inside the 60min.

[mrholm]

Hi
Yes there's these spikes up around 1,2 TB in size for the logs.... that we need to backup and truncate (and later do a shrink of the file).
We currently see that we move logs out from the serve with around 200 MB/s to our proxy and that is OK to me in transfer speed.
That means we need around 1 hour and 48 minutes to transfer it all.
We from backup team don't know when these restores and large jobs takes place so we can only see when the log jobs fail. So we ask for a logic within Veeam that see that a backup is running and skips the "next" planned backup window and make sure to finish the current running log backups. In an environment with +150 SQL servers and 20 FTE SQL DBA this would be a good improvement.

[HannesK]

Hello,
take your request as counted. I will talk to my colleagues and see what could be possible.

Best regards,
Hannes

[mrholm]

Thanks Hannes