Daily full backup to Tape Job Problem

[Zdministrator]

Hello everybody,
first of all, my main language is german and my english is not that good.

Iam very new with Veeam Backup an having a problem with setting up my Tape Job.

My Goal is to have a full Backup from the latest Backup Chain from all my VMs on Tape, every Day from Monday to Friday.

At the Moment my Tapejob writes on the first day everything on tape (Processed: 1,9 TB), but the second day is seems that only the incrementals will be backuped (Processed: 91,4 GB), but i want to backup everyday everything i need for a full restore on one tape. I hope my description is clear enough. I already try different Settings but with no Success.

Here are my current Backup Settings:

Backup Jobs:
Diskjob1: 8 VMs to Backup Repository1 (Retention: Keep 21 Restore Points on Disk, Backup Mode: Incremetal + Active full Backup "Weekly on Saturday") (Schedule: 20:00 Everyday)
Diskjob2: 15 VMs to Backup Repository2 (Retention: Keep 21 Restore Points on Disk, Backup Mode: Incremetal + Active full Backup "Weekly on Saturday") (Shedule: Runs after Diskjob1)

Tape Job:
Tapejob:
Backupfiles: Backup Repository1 + Backup Repository2
Full Backup: Mediapool "Daily" (Full Backup Schedule: Weekly on MO, DI, MI, DO, FR)
Incremental Backup: Mediapool "Daily"
Options: Eject media upon completion, Advanced: Process latest full Backup chain only
Schedule: Daily 23:45 on these Days (MO, DI, MI, DO, FR)


Mediapool "Daily":
Tapes: 5 (Libary LTO-6 HH)
Media Set: Create new media set for every backup session
Retention: Do not protect data
Options: Use Encryption



So, what iam doing wrong?

[PTide]

Hi and welcome to the community!

My Goal is to have a full Backup from the latest Backup Chain from all my VMs on Tape, every Day from Monday to Friday.

So you want to place fulls to tape only, no increments at all, is that correct? If so then you need to switch to forever-forward incremental mode and disable incremental backups processing in you Tape job settings In order to do that you need to contact technical support team so they can provide you with a hotfix as there is a bug in the current version.

Thank you.

[Zdministrator]

I try to uncheck "Incremental Backup" in Tapejob, the first time was fine but the second day after says "No new files to backup"
Its seems like the "Full Backup" Schedule in Tapejob is ignored!?

Is this whats needs to hotfix?

My backup to disk is "Incremetal + Active full Backup "Weekly on Saturday"

On tape i want allways all i need to have a fullbackup:

(for exaple how i imagine:)
Tapejob MO: Full SA + Incremental MO
Tapejob TU: Full SA + Incremental MO + TU
Tapejob WE: Full SA + Incremental MO + TU + WE
...



So i need a Hotfix?

[PTide]

Is this whats needs to hotfix?

Correct.

So i need a Hotfix?

Actually you can switch your job to forever forward incremental mode and configure everyday virtual full. In this case your tape job will compose a full backup on a daily basis i.e. you will have the following schema:

Tapejob MO: Full MO
Tapejob TU: Full TU
Tapejob WE: Full WE
...

Another option is to switch your backup job to reverse incremental - that will make your tape job to copy the latest version full to tape every day.

[Zdministrator]

Actually you can switch your job to forever forward incremental mode and configure everyday virtual full. In this case your tape job will compose a full backup on a daily basis i.e. you will have the following schema:

Tapejob MO: Full MO
Tapejob TU: Full TU
Tapejob WE: Full WE

So i need to uncheck the "Active Fullbackup every Saturday" in my Diskjob to get the everday Fullbackup in my Tapejob to work?

[PTide]

So i need to uncheck the "Active Fullbackup every Saturday" in my Diskjob to get the everday Fullbackup in my Tapejob to work?

Correct.

[Zdministrator]

Can someone please look at this,
what iam doing wrong?

Screenshot:
https://onedrive.live.com/redir?resid=6 ... hoto%2cjpg

[PTide]

Hi,

So your DiskJob1 creates an incremental backup on Repository1at 20:00 everyday, your DiskJob2 runs right after DiskJob1 and creates incremental backup on Backup Repository2, your TapeJob starts at 23:45 daily and is supposed to create only full backups on tape on weekdays, is that correct? Are you positive that new backup files were present on both repositories right before the Tape Job started? Also have you disabled "Archive incremental backups to tape"?

Thank you.

[Zdministrator]

Yes, thats correct!

The Backup to Disk Jobs are finished and iam disabled incremental Backups to Tape.

Here another Screenshot:
https://onedrive.live.com/redir?resid=6 ... hoto%2cJPG


I installed the latest Update but same Problems, sometimes the Tape-Job runs, sometimes not...

[PTide]

Please make sure that you have "Process latest full Backup chain only" checkbox disabled and if that does not help please open a case and post your case ID here.

Thank you.

[Zdministrator]

Thanks for you reply.

"Process latest full Backup chain only" was enabled!
Its disabled now, i will see whats happening and i will give feedback next week!

[Zdministrator]

I disabled "Process latest full Backup chain only" and the first 4 Backups seems good, but now its the same as before... 4xGood, 1xNothing to Backup, 2xGood, 1xNothing to Backup, 1xGood...

[PTide]

Such behaviour looks like one of already known bugs - please contact support team and post your case ID here.

Thank you.

[aaron@ARB]

And for what its worth, your English is perfectly understandable, well done I know some native English speakers who I can understand less than you

[Zdministrator]

Thank you.

As suggested i open an Support Case:

Case ID 01779070

[Didi7]

We have problems with Virtual Full Backup as well. In one of our VBR-environments, we use an HPE LTO7 single drive (not a library), so that a virtual full backup should be written every work day (MO-FR), but unfortunately sometimes the B2T-job reports the following ...

* No new files to backup per job settings
* 5/11/2018 12:15:28 AM :: Synthesized full backup was not created for 5/11/2018: there are no new restore points between 5/11/2018 and 5/11/2018

I don't know, why the B2T-job doesn't compare between 5/10/2018 and 05/11/2018 in this case, as this jobs runs once per day right after the B2D-job, which creates new restore points every day and therefore new restore points for a virtual full backup should be available, but something is fishy with the B2T virtual full backup mechanism.

We have a daily B2D-job (forever forward incremental with 30 restore points) which starts at 11:00pm (MO-FR), 15 VMs are included and 15 new VIB-files are created every day and the oldest VIB is merged into the VBK file.

The B2T-job has the corresponding repository used in the B2D job selected, a full backup (virtual full backup) selected and a deselected 'Archive incremental backups to tape' and runs after the aforementioned B2D-job, which creates 15 new VIB-files from 15 VMs every day (MO-FR), but sometimes or let's say every 2 or 3 days the B2T-job reports the upper mentioned errors and the job is quit with a status => Success: 1 and tape is also not ejected, though it's selected in the options of the B2T-job to eject media upon job completion, whether it makes sense or not in this situation, but tape media is changed everyday here.

Any idea what's wrong with the virtual full backup, reporting 'there are no new restore points between 5/11/2018 and 5/11/2018' instead of 5/10/2017 and 5/11/2018 ?

Best regards,
Didi7

[Didi7]

Any ideas?

Is a HotFix necessary to fix the issues with Virtual Full Backup? We use VBR 9.5 Update 3.

Regards,
Didi7

[Didi7]

Hello

let me try to clarify my problem with 'virtual full backup' again ...

First of all, we have a B2D-job (forever forward incremental) with 15 VMs, one of the VMs is a Windows Domain Controller, so every week day (MO-FR) incremental data is written to the B2D-repository, usually a few GB's of data in the range of 50-100GB per day, so in fact, there is changed data every week day (MO-FR).

The B2T-job runs after the B2D-job (MO-FR), selected is the B2D-repository used for the 15 VMs from the B2D-job, virtual full backup is selected, incremental backup is not selected, process latest full backup chain is selected in Advanced Settings under Options in the B2T-job.

Last night, the following log entries can be seen in the History of the B2T-job:

5/17/2018 12:52:38 AM :: Building source backup files list started at 5/17/2018 12:52:38 AM
5/17/2018 12:52:39 AM :: No restore point for 5/17/2018 is available for virtual full backup, using previous restore point for 5/16/2018
5/17/2018 12:52:40 AM :: 15 synthetic full backups will be placed into the media set
5/17/2018 12:52:41 AM :: Source backup files detected. VBK: 0, VBK map: 15, VIB: 0

Everything went fine and around 3,4TB of data was written to the LTO7 tape media!

A day before, the following log entries can be seen in the History of the B2T-job:

5/16/2018 12:54:53 AM :: Building source backup files list started at 5/16/2018 12:54:53 AM
5/16/2018 12:54:53 AM :: Synthesized full backup was not created for 5/16/2018: there are no new restore points between 5/16/2018 and 5/16/2018
5/16/2018 12:54:54 AM :: Source backup files detected. VBK: 0, VBK map: 0, VIB: 0

This happens randomly (MO-FR). Sometimes it works, sometimes not.

Something is fishy here.

Is this faulty behaviour already known? Is a fix available?

Veeam B&R 9.5 U3 is used in this environment.

Regards,
Didi7

[PTide]

Didi7,

I apologize for missing your posts.

Is there any chance that the B2D job was performing backup chain transformation when the tape job was running?

Thanks

[Didi7]

Hello PTide,

what do you mean by backup chain transformation? This happens during night hours, so that's hard to say for me atm. If you mean the injection of the last *.vib into the *.vbk of each VM, this happens at the end of B2D-job and is reported in the log ...

5/16/2018 11:23:40 PM :: Full backup file merge completed successfully
5/17/2018 12:52:10 AM :: Load: Source 97% > Proxy 25% > Network 11% > Target 0%
5/17/2018 12:52:10 AM :: Primary bottleneck: Source
5/17/2018 12:52:10 AM :: Job finished at 5/17/2018 12:52:10 AM

The B2T-job starts right after the B2D-job at ...
5/17/2018 12:52:35 AM :: Job started at 5/17/2018 12:52:33 AM

A day before, the following can be seen in the History log ...

5/15/2018 11:25:55 PM :: Full backup file merge completed successfully
5/16/2018 12:54:32 AM :: Load: Source 97% > Proxy 21% > Network 10% > Target 0%
5/16/2018 12:54:32 AM :: Primary bottleneck: Source
5/16/2018 12:54:32 AM :: Job finished at 5/16/2018 12:54:32 AM

The B2T-job starts right after the B2D-job at ...
5/16/2018 12:54:50 AM :: Job started at 5/16/2018 12:54:47 AM
5/16/2018 12:54:50 AM :: Processing tasks
5/16/2018 12:54:54 AM :: No new files to backup per job settings
5/16/2018 12:54:54 AM :: Job finished at 5/16/2018 12:54:54 AM

Btw, did you notice that a B2T-Job that runs right after the B2D-job (via Schedule 'After this job'), reports the B2D-jobname under BACKUP column instead the B2T-jobname, the ACTION column corresponds to the B2T-job, strange.

Regards,
Didi7

[lyapkost]

Hi. Could you please also clarify if your source B2D job is a Backup Copy job? And one more, you say the problem happens randomly, is it really so? Doesn't it appears every other day (MO, WE, FR, for example)?

[Didi7]

Hello lyapkost,

no the source backup job is a standard B2D-job (not a Backup Copy job) and appears as Type 'VMware Backup' under Jobs => Backup selecting 'Home' on the left pane. The B2T job (Type => Backup to Tape) has 'After this job' selected in Schedule and runs right after the B2D-job.

Here are the results of the last 4 weeks regarding worked or didn't work (when no changes are found or tape media was not inserted) ...

5/22/2018 1:06:52 AM => didn't work because no tape media was inserted (TUE night)

5/19/2018 1:24:05 AM => worked and 3,4TB was written to LTO7-media (SAT night)
5/18/2018 12:46:18 AM => didn't work because no new files were found (FRI night)
5/17/2018 12:52:35 AM => worked and 3,4TB was written to LTO7-media (THU night)
5/16/2018 12:54:50 AM => didn't work because no new files were found (WED night)
5/15/2018 1:04:36 AM => worked and 3,4TB was written to LTO7-media (TUE night)

5/12/2018 12:55:48 AM => didn't work because no tape media was inserted (SAT night)
5/11/2018 12:15:24 AM => didn't work because no new files were found (FRI night)
5/10/2018 12:50:35 AM => worked and 3,4TB was written to LTO7-media (THU night)
5/9/2018 12:48:42 AM => didn't work because no new files were found (WED night)
5/8/2018 1:09:41 AM => worked and 3,4TB was written to LTO7-media (TUE night)

5/5/2018 1:44:49 AM => didn't work because no new files were found (SAT night)
5/4/2018 12:46:53 AM => worked and 3,4TB was written to LTO7-media (FRI night)
5/3/2018 12:49:02 AM => didn't work because no tape media was inserted (THU night)
5/2/2018 12:45:34 AM => didn't work because no new files were found (WED night)
5/1/2018 1:10:01 AM => worked and 3,4TB was written to LTO7-media (TUE night)

4/28/2018 1:21:46 AM => worked and 3,4TB was written to LTO7-media (SAT night)
4/26/2018 11:50:31 PM => worked and 3,4TB was written to LTO7-media (THU evening)
4/25/2018 11:53:26 PM => worked and 3,4TB was written to LTO7-media (WED evening)
4/25/2018 12:03:46 AM => didn't work because no new files were found (WED night)
4/24/2018 12:20:53 AM => worked and 3,4TB was written to LTO7-media (TUE night)

Best regards,
Didi7

[lyapkost]

That's a known issue discussed earlier in this forum and this behavior is planned to be improved in future releases. It takes place when tape virtual full is scheduled for every day and source job starst before 12 PM and finishes the next day (your case) or source job is a Backup Copy job.

There are two workarounds: either reschedule your B2D job so that it finishes the same day when starts or split tape jobs into two with schedule of virtual full 1) MON, WED, FRI & 2) TUE, THU, SAT.

[Didi7]

Hello again,

is it really possible to create one B2T-jobs, which starts via ‚After this job‘ in Schedule option and selecting MON, WED, FRI for virtual full backup and another B2T-job, which starts via ‚After this job‘ in Schedule option and selecting TUE and THU for virtual full backup?

Both B2T-jobs start after the same B2D source job?

Planned to be improved in future releases? That‘s a very nasty bug and not a feature request.

Regards,
Didi7

[lyapkost]

is it really possible to create one B2T-jobs, which starts via ‚After this job‘ in Schedule option and selecting MON, WED, FRI for virtual full backup and another B2T-job, which starts via ‚After this job‘ in Schedule option and selecting TUE and THU for virtual full backup?

It is.

Both B2T-jobs start after the same B2D source job?

They do. But since one of these jobs always doesn't have virtual full scheduled for today, it will finish immediately, the other one will start creating virtual full.

[horaceho]

Hi all,

I had the same issue here and use the work around like lyapkost talk , I create 2 B2T job : 1 is at 23:30 daily for the backup job from 12:00pm to 22:00 (because I need to confirm all backup job can completed before 23:30) and another B2T job start at 05:00 for backup job from 00:00 to 04:00 . because I off-site tapes daily 10:00. so far it is normal over 1 month.

but finally I modify all the backup job use reverse inc from this week , because it still have 1 -2 times some of backup cannot create virtual full to backup tape and I need to start a active full manually . it's not stable for me.