Help with Backup Copy Job Settings

[markp]

Hello,

I would like to set up a backup copy job that keeps the following restore points:

Days: Last 30
Weeks: Last 8 Before the 30 Days
Months: Last 9 before the 8 weeks
Quarters: Last 8 before the 9 months

Most Recent -> Oldest

D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D W W W W W W W W M M M M M M M M M Q Q Q Q Q Q Q Q

I have set the following:
Restore Points: 55 (30Days + 8 Weeks + 9 months + 8 Quarters)
Archival: 8 Weekly, 9 Monthly, 8 quarterly.

Is this correct?

In this scenario when would the first weekly Full be made? I currently have 1 full vbk with 11 daily vibs. I am wondering why the weekly has not run yet. Does it wait until i pass the 55th restore point to do a weekly?

Thanks,

[veremin]

Hi, Mark.

You don’t have to take into account GFS restore points, since they will be created from the existing backup chain. So, the 30 as number of simple retention policy will be enough.

As to weekly restore point creation, you will have to wait until the backup chain exceeds set limit (30) and, then, till the full backup file reaches the day when the weekly backup is scheduled.

More information regarding how GFS rotation scheme works can be found in the corresponding User Guide (p.114); so, kindly take a look.

Thanks.

[markp]

Thanks for the clear answer.

Do I not need to worry about the number of restore ponits before creation of the Full as in standard backups. Generally, I would not want so many incrementals between a Full buckup creation?

[veremin]

There is no active full backup concept in case of backup copy job. Though, from time to time you might want to run compact full backup in order to get full backup file defragmented, reduce its size and increase the speed of writing and reading to/from it.

Thanks.

[markp]

Thanks for your help.

[veremin]

You’re welcome. Feel free to ask, should any additional clarification be needed. Thanks.

[Kantholy]

Hi there,

I have a problem with a Veeam Backup Copy Job I created for the longtime archival: for some reason, the GFS scheme is not applying to the job... I read the documentation and created a backup and backup copy job for purpuse:

Backup Job: Name: "SCH-FS": 1 VM "sch-fs", one a day, 5.00 am
Backup Copy Job: Name: "ARCHIVE_SCH-FS": copies the backup job "SCH-FS" to a backuprepo, GFS enabled, keep 4 weekly and 12 monthly backups, retention period: 30days
---
as far as I understand the documentation, this backup job copy should keep 30 days of my daily backups in "incremental" stage and creates a weekly full backup once a week...

Now I waited a whole month so I can see wheter its working or not... but unfortunately is looks like that the GFS scheme is not applied for some reason: the folder ARCHIVE_SCH-FS on the filesystem only contains 1 full + 29 incrementals...


the following two pdfs show all settings for the Backup + Backup Copy Job to fully see my configuration of those jobs.
Configuration of Backupjob SCH-FS
Configuration of BackupJobCopy ARCHIVE_SCH-FS

can someone please help me? I don't get what I did wrong

thanks in advance

[foggy]

Tobias, with current implementation of GFS, the first weekly full will be created after reaching the specified number of restore points to keep. Until this happens, we will still maintain the required number of weekly/monthly fulls as per your GFS retention settings, just in the incremental form. Please see the topic above, thanks.

[tigershark]

Hi guys,
Sorry if this question has been asked before.

I've set up a Backup Copy Job 2013-09-17 and it have been running since that. I thought that it would be nice to keep 4 weeks, 12 months, 4 quarterly and 10 years just for archival purposes.

So I choosed the following settings:


But every night all the restore points keeps being pushed forward except the first one created on 2013-09-18.


Do I need to set Restore points to keep to 30 (weeks+monthly+quarterly+yearly) instead of 7? I thought that restore points was only for the incremental backups. Or am I doing something else wrong? Looking forwards for some expertise help

[veremin]

I’m wondering whether you’ve specified these settings originally, or you’ve added GFS rotation scheme a lit bit later. Also, I recommend you waiting till the existing full backup is pushed to the Friday and see whether weekly-worth restore point will be created or not.

Thanks.

[nunciate]

I am still a bit confused on how the GFS and Restore Points work together. Anyone have a link to something that give more detail on this.
I do my daily backups to disk with a weekly synthetic full and revers incrementals on Fridays. My restore points are set to 25. I need to have 30 days of backups available on disk.
I want to do a backup copy job to send this data to an offsite server. I want the exact same 30 days of data. I don't want to archive for longer. I use tapes to archive for longer periods.
So do I just set my restore points for the backup copy job to 25 as well and that should do it?

[veremin]

I want the exact same 30 days of data.

So do I just set my restore points for the backup copy job to 25 as well and that should do it?

If you want to have 30-day worth backup data, then, disable GFS rotation scheme and set 30 as a number of restore points backup copy job should keep. Thanks.

[tigershark]

I’m wondering whether you’ve specified these settings originally, or you’ve added GFS rotation scheme a lit bit later. Also, I recommend you waiting till the existing full backup is pushed to the Friday and see whether weekly-worth restore point will be created or not.

Thanks.

I've specified the GFS when I created the job. So since Friday already has occured three times (Sept 20th, Sept 27th & Oct 4th) since I created the job there should be a couple of weekly there right?
Do the settings needs to "match" the ones on the original backup job? On that job I use incremental with enable synthetic fulls (forever-incremental) on Friday, could that be a problem?
If that's the case, how would I set up the jobs to work as I intended it to do?

[veremin]

No, the GFS settings of backup copy job have nothing to do with those of backup job.

Your settings look good as to me. So, I suggest that you open a ticket with our support team and let them investigate directly, as the whole situation seems to be unexpected.

Thanks.

[danielchung]

I have a backup copy job with below settings:

1 day interval at 8:30pm
10 restore points
keep 2 weekly restore points at Friday 8:30pm

I expected the job would generate the latest weekly full backup every Friday but it did not. Now I have backups below:

- Backup chains starting with full backup as of Oct 4
- Weekly backup as of Sep 27
- Weekly backup as of Oct 4

Weekly full as of Oct 11 was expected but it is missing. It seems transforming the weekly full only when the full backup in the backup chains reached Friday. Do I wrongly understand the weekly full backup schedule?

[veremin]

Hi, Daniel.

I believe you will have to wait until the existing full backup file is moved to the day when the weekly backup is scheduled (Friday). Taking into account your retention settings, it should happen on Monday, 21st. Thanks.

[danielchung]

So the weekly backup schedule is correlated to the number of restore points in simple retention policy?

[veremin]

As mentioned, weekly backups are created from a full backup of the regular backup cycle. Therefore, in order for weekly backup to be created, the backup chain should reach specified number of restore points (10, in your case) and existing full backup should reach the day when the weekly backup is scheduled (21st, October).

Also, it might be worth reviewing corresponding section at our Help Center.

Thanks.

[danielchung]

Sorry to bring the old thread up again.

After a period of testing, the latest backup in backup copy job would never be a full backup (except for the initial run). It contains at least 2 restore points (full + incremental). Following up my situation, is there anyway to generate or transform the last backup from backup copied files into a vbk file so that I can offload it to tape?

[Gostev]

the latest backup in backup copy job would never be a full backup (except for the initial run)

Why? Sure it will. Every restore point will become the full backup sooner or later. If you maintain 10 restore points, then this will happen in 10 days since the given restore point has been created.

[danielchung]

What I mean is that the last created backup file is always a vib file, rather than a vbk file. I'd like to offload the last backup as a full backup to tape, instead of offloading a 10-day old full backup.
If backup copy job can either keep only 1 restore point or do something like reverse incremental will make it more simple.

[veremin]

Hi, Daniel,

You can either transfer to the tapes 10-day old full backup or try some creative workaround. In case of workaround, you can try the following:

1) Create new backup job and map to existing backup copy job.
2) Create new backup copy job and choose the newly created backup job as a source for it.
3) Run the newly created backup copy job.

After successful run, there will be a .vbk file containing the latest state.

In fact, I haven't tried it myself, but, based on my understanding, it should work fine.

Thanks.

[danielchung]

I've already tried these steps for creating a seed for backup copy job so I knew it worked.
However, the offloading task has to be done regularly, say weekly. For one backup job, I have to create 1 backup copy job for offsite, 1 extra backup job for dummy mapping, 1 extra backup copy job for vbk creation. I'm managing 1x backup jobs for different sites, this workaround involves manual work and creation of 2 extra jobs for each site. I don't think it's practical to me.

Frankly, the backup copy feature should be enhanced with function to support either:
1. Copy from another backup copy
2. Reverse incremental
3. Regular creation of full backup vbk

I think it makes sense to keep the latest backup as a full backup offsite or in DR location.

[foggy]

Thanks for your feedback. To justify the forward incremental "nature" of the backup copy jobs, I would mention that reversed incremental mode puts even more load on the target storage due to more random I/O pattern. Besides, there are a couple of reasons more that govern this kind of design.