Archiving Restore Points

[Helferlein]

Hi all.

I have a fileserver that I actually would like to backup every two hours.
I set all after 30 days shall go into the azure object storage.

Now it seems that when there is a big... (well 10GB) restore point to archive it take like 3 hours to upload this!
Transferring with 969KB/s.

And while the backup job is doing it's post-processing and archives the restorepoints the job does not start every two hours anymore...
It will start when it's done archiving.

That is not really cool... Wouldn't it be better to abstract the archive process from the backup process?
Just like you do it with SOBR tiering or even with the backup copy jobs?

[Dima P.]

Hello Peter,

Now it seems that when there is a big... (well 10GB) restore point to archive it take like 3 hours to upload this! Transferring with 969KB/s.

Just to check - such transferred rate is expected in your environment, right?

Wouldn't it be better to abstract the archive process from the backup process? Just like you do it with SOBR tiering or even with the backup copy jobs?

Will discuss with the RnD folks if it's possible, thanks for sharing!

[Helferlein]

Dear Dima.

This transfer rate is absolutely slow. We have two 1Gbit/s symmetric enterprise internet connection. So just based on the upload speed this should go much faster.

If it would go faster, I think the archiving would not be a big problem.

But a month ago there I had every hour backups. So there are daily between 18-24 restore points with sizes of around 10-20GB to archive now 30 day later.
Tonight it did work on around 24 restore points like 14 hours. Transfer speed between 700KB/s up to 2MB/s.
The last transforming backup process took 5 hours.

But I think I found the problem in my case:
The thing is, that I have set only one azure blob storage archive storage for two of our biggest fileservers.
And currently the other job is running since days a repair job of missing data.

In this case I think it's better that I stop this repair job and change that job to an other separate azure blob storage so they would not interfear eachother.

[Helferlein]

Well the above mentioned repair job was not the problem. Nor that both jobs did go into the same Blob storage.
The problem persists.

And what I see is, that Veeam just starts archiving jobs with the latest job per day... So I have several runs of that job during the day (running actually every hour if possible). And with all of them Veeam does not archive restore points.
It starts with the last job of the day. All restore points from last month at once. Since in my case this takes sometimes 9-10 hours the first job that runs next day again is almost at noon time and I do not have restore points in the morning.

It seems that the proxy is the bottleneck. I will try to add another proxy.

But still... Why Veeam does not process archives after every running job some instead of all together when the new day did start?
Or I think, that it would be really the best to abstract backupjobs from archiving jobs.