Orphaned and duplicate files from multiple jobs

[seapro220]

Hello all.
I originally created a 'production' job which included about 22 servers. Last week I removed about 1/2 of the servers, and put them into a 'job' of their own. I've also had to run a 'full backup' on one of the jobs, which has generated 2 vbk files, instead of the usual weekly (forward incremental) vbk file. As time is passsing, I've started collecting vrb/vib files on both jobs as well as having a couple of 'duplicate' vbk files out there. Space isn't an issue at this time on my NAS, but I dont want to let this 'go on forever' .. know what I mean. What I believe is happening is this, please confirm -

1. original job has 'remenants' of all of the restore points (jobs are set up as 30-day retention) from all the servers. The night i changed/removed/split the servers - then the vib/vrb files decreased on the original job - and the weekly VBK should have also reduced/changed because of the different # of servers in it.
2. The new job has run processing 1/2 of the servers, and it's 'info' should be correct with only 1/2 of the servers in it. It's weekly VBK file should also only reflect the 'new' servers pertaining to those jobs.

Sound right - ??

Now, how do i know 'when' I can delete the 'manual full backup vbk' files, and if I have to restore, how long do I know which restore file - has which data in it that I need? Presumably the jobs will 'trim' themselves out during the normal 30-save process...but, do i have to wait that long? Should I archive that data off, change the job from 30-day to 1 day, and force a 'full' backup to recreate/start over? This doesn't seem to make much sense and is, as you can image potentially stricken with 'murphys law' .


mark

[Vitaliy S.]

Hello Mark,

1. Yes, that's correct.
2. That's also correct.

For situations like that - when VMs are no longer backed up by the job we have another setting called "Deleted VMs" retention period. Take a look at this topic which covers everything in details: Deleted VMs Retention Period

If you have any further questions, please let me know. Thank you!

[seapro220]

The post is a little confusing, but let me see if I understand ...

1. A job is set for 30 days of retention, reversed incrementals and every Sat night it 'rolls up' the files into the new 'full' vbk file. In this manner, I have a weekly 'full' backup, and 30 days of restore points - right?
2. VM retention is set for 15 days. During this 15-day timeline, and servers added/removed will automatically drop/add out of the job.

I think this is what the rule is stating -

If a pictular vm, part of this job - hasn't been backed up during 15-days, then it's automatically trimmed from the backups. So, basically, keep a 'log' for 15 days of history on all the servers in this job and if 1 drops out - start counting 15 days and automatically prune it's data (restore and restore points). If a server is added during this process, then start the clock on 15 days before you do something.

If this is true, how can you determine how much 'data' will be trimmed from the backups or vrb/vib files - to know if you should expect to 'gain' x-%' of space back? This isn't that critical in determining as I image it's hard to determine (30 gig server image, 40-meg average changes over 15 days) how much you actually retrieve back - but just curious.

mark

[foggy]

1. A job is set for 30 days of retention, reversed incrementals and every Sat night it 'rolls up' the files into the new 'full' vbk file. In this manner, I have a weekly 'full' backup, and 30 days of restore points - right?

In the initial post you've stated that you use forward incremental mode. Is this the case?

2. VM retention is set for 15 days. During this 15-day timeline, and servers added/removed will automatically drop/add out of the job.

Please review this post for better understanding of how Deleted VMs retention setting works and what it refers to.

So answering your question about space reclaim, no data is "trimmed" from the already produced forward incremental backups without transform before retention comes into effect and deletes those old backups that contain the removed VMs. Just newly created incrementals and fulls will become smaller as will not contain removed VMs data.

Things differ in case of reversed incremental though.

[seapro220]

Yes, the jobs are set as 'forward-incrementals'.
BTW - I just looked at my 'stats' from last nights job, the one that was modified and I see a 'note' in there about "XX-server is no longer procssed by this job, verify..". I presume this is part of the notification that 'that' pictuliar server isn't being backed up by that job anymore. Good.
On a side note - pertaining to this same job - I have one of my servers in this job that is still 'preparing guest for hot backup' after 12+ hours. This shouldn't have taken (at most) 12 minutes - not hours, to process and complete.

Thoughts?

[foggy]

On a side note - pertaining to this same job - I have one of my servers in this job that is still 'preparing guest for hot backup' after 12+ hours. This shouldn't have taken (at most) 12 minutes - not hours, to process and complete.

Mark, job logs review is always required for troubleshooting technical issues like that, please contact our support team directly. And also in order not to derail this thread, it is advised to create a separate topic for any unrelated questions/issues. Thanks.