1

I can’t seem to get the right Retention settings for a scheduled Archive job on my Unitrends appliance. The scheduled Archive doesn’t delete old backups to make room for the new and fails. I’ve been manually deleting the archives on my NAS and then deleting the Archive Set under the Status tab.

Attached is a screenshot of my scheduled Archive.

What am I doing wrong?

unitrends.jpg

5 Spice ups
2

I should note, my oldest archives have a Retention Period of 4 Months.

I changed to 0 Days a few months back, so there are a few backups to delete before I get to the first 0 day.

Is this the issue?

1 Spice up
3

Why is your retention set to 0? If you are archiving, I think that number should represent the number of archives you wish to keep, and the oldest gets purged.

1 Spice up
4

I think you are on the right track, I would try maybe setting that number to 1 or two and then seeing if it will delete the oldest archive once the schedule is in effect.

5

Here’s a full explanation:

http://www.unitrends.com/documents/administrators-guide/user_manual/archiving/archive_settings.htm

1 Spice up
6

There’s a ton of info on their KB site that you may want to dig through a little:

http://support.unitrends.com/ikm/categories.php?categoryid=18

1 Spice up
7

Eric - Have you gotten this resolved yet?

8

Hi! I am also having the same issue as PokerMunkee but with archiving to eSATA disks instead of NAS. I have a case open with support regarding this but haven’t really received a clear answer as to whether this is a known bug. I reported multiple little issues with archiving in my ticket and the support tech said maybe some of the quirks I reported would be fixed with an upgrade to 7.3. It was not really clear which ones (if any) were actually going to be fixed. Regardless, updating is not an option for me right now because I am a Vault2Cloud customer and our updates are managed by the cloud team.

Here is my setup. I have 4 sets of eSATA disks that I rotate offsite each week. My two archive jobs run weekly on Sunday and both have the purge option checked and the retention period set to 18 days. Each week I get a set of disks back from offsite storage to use for the next archive job and the sets on those disks are about 21 days old. However if I run my jobs without first manually preparing the disk to erase the data, I often run out of space during my second archive job due to the disks being out of space.

My understanding of how purge works is that a job with the purge option checked should purge any sets on the disk older than 18 days with my configuration. However that is definitely not happening. I just put in my disk set that came back today. I had two archive sets on the disks from 12/16, which is 22 days ago. As a test I ran a small archive job with the purge option checked and 0 days retention and after the job I looked at the archive sets on the disk set. I then had 3 archive sets on the disks, the two from 12/16 are there and the third one (1/7) that just completed. Furthermore I ran another test, this time with the overwrite option set. After that job completed the only archive set on my disks was the job that had just run, which confirms that the system recognized all of the prior sets as being past their retention period. So while overwrite option seems to work, the purge option definitely does not.

I am still waiting on support to be more clear with me, but I was wondering if you can tell us if s this a problem that is documented and being worked on currently. And if so, is there a projected time when it should be fixed?

Thanks!

9

Pat,

have you looked through any of the links I provided to PokerMunkee?

10

Pat - Thanks for choosing Unitrends!

Have you been experiencing these issues for a specific period of time, or forever? Also, if you can forward over a case number or asset tag number, I can find out what things that you’ve submitted are on the roadmap, or have already been addressed, as well as ensure that someone from Support connects with you ASAP.

11

Bill Kindle - Yes I read through the administrators guide link before posting, though I have looked at that before. It still reads to me the way I expressed it in my previous post, which is purging happens by date not by # of archives to retain.

Katie - my support case # is 113219. I have experienced this for as long as I have been doing archiving, circa July 2013.

Thanks for your attention to this!

12

Pat - Always happy to help! I’ve engaged our Customer Support Management team, and they are looking into this for you.

13

I just got off the phone with a support technician who clarified things for me. My main takeaway is that the information in the administrators guide is a little misleading as it does not work the way it implies. And apparently PokerMunkee and I are not the first people to be thrown by this. He did say they are trying to get it better documented so folks are not so confused in the future.

It sounds like the archiving to disk and purging is a bit more complex than you’d think. It is not like an old school backup program like Backup Exec where you tell it when you want to mark a job as expired/erasable after which it will be overwritten. The archive operation sounds like it is geared toward allowing you to fit as much duplicate data on disk as possible. As with backup jobs on the appliance, it requires multiple copies(2 or 3 - he was not sure) to deduplicate the backup sets. Furthermore the system will not purge any sets off the archive disk until it is able to deduplicate the data. Effectively this means it will never purge until you are up to your 2nd or 3rd archive set of the same data. (This may explain Bill Kindle’s take that the retention is based on the number of sets you want to keep.)

This also means that you have to have 2-3 times more space on your archive media than the amount of data you are archiving in order to hit the deduplication target. Maybe this is what PokerMunkee is running in to? Perhaps adding more space to your NAS would help.

For my purposes, since I am using removable eSATA disks I not worried about deduplicating data. My goal is to have multiple disk sets offsite for disaster recovery purposes. I will just overwrite my disks each time. I am going to try setting the first of my two archiving jobs to “overwrite” so that it will automatically erase the disk before the job starts. The support guy said just to be safe, I may want to set my second job with a retention policy of 0 days to make sure it also erases.

Also, apparently nothing in release 7.3 will change/fix anything with archiving.

Again, Katie, thanks very much for your help!

14

One last thing… FWIW, I think maybe there should be multiple modes of archiving. For people who want more long term retention space on a NAS, the way they have it set up makes some sense as deduplication lets you get the most bang for your storage buck. However for people who just want discreet copies of backups on different disks (e.g. Week 1 backups, week 2 backups, etc.) it seems like it could be more basic and allow you to delete stuff off disk as it expires.

15

Pat - Thanks for this rundown of your call. I did want to clarify that the more copies that you have of your data, the better that deduplication works. You do only need three copies of the data, as dedupe runs after your next backup has completed since most restores are done from the most recent backup. With that in mind backups 1 & 2 are deduplicated after backup 3 has run. If you don’t have the space to keep 3 copies of your data on your archival media, then you may want to turn off deduplication on the archive target. Does that make sense?

16

Yes, I think so. Are you saying that it would purge expired sets as I was expecting if deduplication is turned off on the archive target? Is that done by unchecking the “compress” check box in the archive job options? Does it matter if your backups on the appliance are already compressed?

I’d have to experiment to see if that works…