Discussions related to exporting backups to tape and backing up directly to tape.
Post Reply
Moebius
Veeam ProPartner
Posts: 206
Liked: 28 times
Joined: Jun 09, 2009 2:48 pm
Full Name: Lucio Mazzi
Location: Reggio Emilia, Italy
Contact:

Understanding GFS jobs

Post by Moebius »

I'm trying to achieve a simple archiving schedule.

All my source backup jobs follow the forward incremental forever method, with different schedules. I have a mix of jobs that run hourly, 4 times a day, daily, weekly, monthly and some are not scheduled, i.e. run only under request.
Some of these jobs were converted recently from reverse-incremental to forward-incremental forever.

What I want is to weekly export a full backup of all my vms, that is, every vm should have its most recent full backup synthesized on tape.
Since I have an IBM TS3200library with two drives, and wanting to use both simultaneously, I set up two GFS media pools with the following settings:
Weekly media set overwrite protection period: 4 weeks
Use any available media; do not append; export to vault XXXX;
Monthly, quarterly and yearly media sets disabled.

Then, I set up two tape backup jobs using these two GFS media pools, balancing the source backup jobs more or less equally between the two, and scheduling both jobs on Saturdays.

The first run of these jobs, however, gave some unexpected results.
Most source jobs were processed correctly. A few, however, were skipped and I don't understand why.
One source job showed this message: "No restore point for 26/03/2016 is available for virtual full backup, using previous restore point for 25/03/2016" - and then did nothing.
Two showed "Source backup files detected. VBK: 0, VBK map: 0, VIB: 0" - and then did nothing.
Another one said "Synthesized full backup was not created for 26/03/2016: there are no new restore points between 30/07/2015 and 26/03/2016" and did not backup anything. 30/07/2015 is the date of the only restore point for this job.

I re-started the jobs manually today hoping to get the missed jobs to tape, but nothing was done.

I guess I'm missing something on how the GFS tape jobs work.
veremin
Product Manager
Posts: 20271
Liked: 2252 times
Joined: Oct 26, 2012 3:28 pm
Full Name: Vladimir Eremin
Contact:

Re: Understanding GFS jobs

Post by veremin »

Are you on Update 1 already, because there are some inconsistencies in GFS candidate selection logic in version 9 GA? Thanks.
Moebius
Veeam ProPartner
Posts: 206
Liked: 28 times
Joined: Jun 09, 2009 2:48 pm
Full Name: Lucio Mazzi
Location: Reggio Emilia, Italy
Contact:

Re: Understanding GFS jobs

Post by Moebius »

Well, the first GFS run was on version 9 GA. I have then updated to U1 but the second run is scheduled for next Saturday. I'll see what happens.
Should I recreate the jobs and/or the media pools that were created with v9 GA or can I keep them as they are?
Dima P.
Product Manager
Posts: 14396
Liked: 1568 times
Joined: Feb 04, 2013 2:07 pm
Full Name: Dmitry Popov
Location: Prague
Contact:

Re: Understanding GFS jobs

Post by Dima P. »

Lucio,

Indeed, there was some issues with the GFS candidate selection logic but per QA it should be resolved in Update 1. Could you check how it behaves after upgrade and let us know? Thanks.
WinstonWolf
Veteran
Posts: 284
Liked: 11 times
Joined: Jan 06, 2011 8:33 am
Contact:

[MERGED] : No GFS candidate for Monthly media set

Post by WinstonWolf »

Hello ,

I configured an GFS Media Pool .
In this Pool there was already an older Mediaset ( cataloged ) .
Yesterday at 00.00 the Monthly Tape Job for March started but the following Message appears and the Job is in the Morning finished in 30 Seconds:

31.03.2016 00:00:14 :: Building source backup files list started at 31.03.2016 00:00
31.03.2016 00:00:15 :: No GFS candidate for Monthly media set was found for today, waiting
31.03.2016 00:00:15 :: Source backup files detected. VBK: 0, VBK map: 0, VIB: 0

What Means this, what goes wrong , why the Job cant find Files ?

Thanks.
veremin
Product Manager
Posts: 20271
Liked: 2252 times
Joined: Oct 26, 2012 3:28 pm
Full Name: Vladimir Eremin
Contact:

Re: Understanding GFS jobs

Post by veremin »

Kindly, install Update 1 and see whether it fixes your issues. Thanks.
WinstonWolf
Veteran
Posts: 284
Liked: 11 times
Joined: Jan 06, 2011 8:33 am
Contact:

Re: Understanding GFS jobs

Post by WinstonWolf »

I have allready installed Update 1 , before Update 1 it works on this Job .
Another GFS Job works fine.

Please Veeam fin an Solution . Is this correct that it is an Bug ?!?!

What is the Statement from Veeam Side ?
Dima P.
Product Manager
Posts: 14396
Liked: 1568 times
Joined: Feb 04, 2013 2:07 pm
Full Name: Dmitry Popov
Location: Prague
Contact:

Re: Understanding GFS jobs

Post by Dima P. »

Hi,
31.03.2016 00:00:15 :: No GFS candidate for Monthly media set was found for today, waiting
Can you confirm that you have a restore point for the set date created by the source job? If the source job is still running, GFS job may wait till the backup is completed.
WinstonWolf
Veteran
Posts: 284
Liked: 11 times
Joined: Jan 06, 2011 8:33 am
Contact:

Re: Understanding GFS jobs

Post by WinstonWolf »

The Source VM Backup Job is started at 9 pm , it is finished at 0:30 Am next day . The GFS Job is not shedulable , it starts only at 0:00 and it is also not possible to set an Time for waiting if a nother job is running , like at the normal Tape Job .
Therefore for me the GFS Job / Pool ist totaly nonsens .
veremin
Product Manager
Posts: 20271
Liked: 2252 times
Joined: Oct 26, 2012 3:28 pm
Full Name: Vladimir Eremin
Contact:

Re: Understanding GFS jobs

Post by veremin »

The Source VM Backup Job is started at 9 pm , it is finished at 0:30 Am next day.
And on which day did GFS job take place? On the first or on the second one? Just trying to understand your situation better. Thanks.
Moebius
Veeam ProPartner
Posts: 206
Liked: 28 times
Joined: Jun 09, 2009 2:48 pm
Full Name: Lucio Mazzi
Location: Reggio Emilia, Italy
Contact:

Re: Understanding GFS jobs

Post by Moebius »

Dima P. wrote:Lucio,

Indeed, there was some issues with the GFS candidate selection logic but per QA it should be resolved in Update 1. Could you check how it behaves after upgrade and let us know? Thanks.
I upgraded to Update 1. Some issues seems to have disappeared but some others don't.

- An entire source job was skipped for no apparent reason.
- Another source job was skipped. This is a backup job without schedule that is only run manually and has only one restore point.
- Another source job threw an error but seems to complete successully.
- Some VMs are shown with a wrong number of restore points in the Backups->Tape window.
- One VM has 4 full backups sent to tape within the same tape backup session.

Case # 01753041.
Dima P.
Product Manager
Posts: 14396
Liked: 1568 times
Joined: Feb 04, 2013 2:07 pm
Full Name: Dmitry Popov
Location: Prague
Contact:

Re: Understanding GFS jobs

Post by Dima P. »

Hi Lucio,

Thanks. I noticed that support team is going to setup a WebEx meeting. Let’s wait for the results.
Moebius
Veeam ProPartner
Posts: 206
Liked: 28 times
Joined: Jun 09, 2009 2:48 pm
Full Name: Lucio Mazzi
Location: Reggio Emilia, Italy
Contact:

Re: Understanding GFS jobs

Post by Moebius »

Moebius wrote:I guess I'm missing something on how the GFS tape jobs work.
Well, after wasting some time with support tier 1, I asked to escalate my case and tier 2 quickly found some reasons why some of my source jobs were skipped.

What support told me is very interesting, and I believe this has not been mentioned in the user guides or documented anywhere. So here it is.
Veeam support wrote: There are a few things of note:

- Veeam will only search from the restore points that were created after you've linked (added) the source jobs to the GFS tape job. So yes you have to run them at least once again after adding -- this may be a bit confusing.

- If you disable a GFS Tape and then enable it, we'll only look for restore points that were created after you've enabled the job. Here we assume that you wanted to skip a period and don't want any backups from the period that the job was disabled.

- Otherwise the latest restore point is taken and copied/synthesized to tape, if it wasn't already copied before.
This kinda explains why my original strategy won't work. I thought I could send a full copy of all my source jobs to tape with a weekly GFS - even those source jobs that are scheduled to run once a month or on demand. The last sentence says they will be copied only once after their last run, and will be skipped in the following tape backup sessions.
As I told support, I fail to understand the logic behind this "limitation", but I will have to rethink my strategy. I might have to enable the monthly GFS and schedule regular backups even for the jobs that run only on demand at the moment (eg. VMware templates).
veremin
Product Manager
Posts: 20271
Liked: 2252 times
Joined: Oct 26, 2012 3:28 pm
Full Name: Vladimir Eremin
Contact:

Re: Understanding GFS jobs

Post by veremin »

Currently, it's by design. If no new restore points have been created by source backup job within an interval (in your case one week), a GFS tape job won't copy a restore point a new, since it has already copied it.

But your question has given us some "food" tor second though about existing logic.

For now you can enable monthly GFS in backup to tape job settings in order for both primary and secondary job schedules to coincide.

Thanks.
Post Reply

Who is online

Users browsing this forum: No registered users and 23 guests