Rotated hard drives support in v8

[Gostev]

That is correct.
Veeam will create the folder.

[zone]

Hi guys,

I'm still having issues after changing to backup copy jobs for daily usb drive rotation. My main problem is with what retention period to set on the backup copy job.
For example; If I have 5 disks and swap them daily monday to friday what retention period would i set to on the backup copy job to hold at least 1 full backup on each disk.

Regards
Adam

[veremin]

You can specify whatever retention you want to. Just assign repository role to rotated media and enable "rotated" option in its settings. VB&R will make sure that at least one .vbk exists on every rotated drive. Thanks.

[zone]

I do have that option set for the removable storage repository. What I am finding is that on the backup copy job I have set "restore points to keep" at 2 due to drive size. Then after a few rotations the drives will end up with 3 - 4 .vbk files which then cause the drive space become full.

[veremin]

What product version you're using? Starting form version 8 there is a corresponding option in the repository settings. Thanks.

[zone]

This is using version 8.0.0.817

Like i said; the repository is set for drive rotation. http://i.gyazo.com/601f9a710b793a3160f6957c1b5acef8.png
Then the backup copy job is set to use this repository however the drive fills up with more restore points than specified under "restore points to keep". In this case it is set to 2. http://i.gyazo.com/2d3cd51bb4320fb25dd232f2ab26becc.png

[foggy]

Basically, with these settings, backup copy job should continue incremental chain on each of the drives, so I recommend contacting support to investigate why full backup is created during each cycle.

[PaulJeM]

Hi Alexander,

We are experiencing similar problems with Backup Copy Jobs, Rotated Media and Incremental Chains. (including Multiple VBKs some days and seemingly successful VIB Incrementals on other days, all from the same Job)

I have created a Support Ticket that is still outstanding from over 2 weeks ago (case 00731632). Each response that we have recived from Veeam Support has stated that Incremental Chains on Rotated Drive is not possible ... but this is not the results we are seeing. We are seeing that its does seem to be possible, if a little flawed!

Our last Veeam Support response stated;

"However this triggered conflict in a Veeam memory, for some reason backup chain was broken and we get mixed files in a repository one for January 20 and the second for January 26. Situation on the print screen was created by mixed of a different restore points for different time shifts, I am not even sure if this combination would be successful for restore.
Currently I see that algorithm which you are trying to use doesn't work like you need. Veeam doesn't support for each drive a single backup chain. It is a by design behaviour. Every time when a new disk inserted Veeam search for the latest restore point in a chain. We do not save for each drive an own chain, we think about all these drives like about one repository."

Obviously we are concerned the reported 'Successful' Backups may not be 'successful for restore' but we are also unsure about what is and what is not possible.

If you could you please clarify what is and is not possible , we would be very grateful?

Regards

Paul

[foggy]

Paul, incremental chains are possible, however there are known issues regarding this functionality (including periodically created full restore points) that will be addressed in patch 2. That said, restores should work fine, if the job reports it was successful.

[joosef619]

Just upgraded to v8 and so far so good but question regarding the repository changes for enabling rotated hard drive support.

Now that we have the option of configuring a repository for rotated media I had removed the previous v7 registry hacks that enabled this:
ForceCreateMissingVBK - For starting the chain anew.
ForceDeleteBackupFiles - For purging broken chain data.

It seems that the new changes in v8 for the rotated media repository are only impacting the creation of a new chain of backups, but not purging the older data. It is removing data but only based on the backup copy job retention assignment.
I did like this automated feature of purging the disk when the disks get swapped on a weekly basis. Wondering if there was something I missed configuration wise, or that I simply need to continue to use the registry hack for now.

Any advice would be appreciated

Joe

[Gostev]

Hi, Joe. In v8, all backup files which are no longer under retention should be purged automatically as you attach an older drive. Please, double check if the backup files you are expecting to be purged are still under job's retention policy. If they are not, please open a support case for investigation. Thanks!

[joosef619]

Thanks Gostev!

Right now all my backup copy jobs are set to only keep 7 restore points.
I rotate the drives once a week, with about 5 weeks worth of rotated media.

As an example I just swapped in a disk for this weeks backup after upgrading.
The backup copy chain has a new full VBK - 3/14
The old chain has 7 restore points - vib and 1 full vbk dates 1/30 thru 2/6.

Checking if that would be considered a broken chain and that data should be purged?
If it should and it isn't I will submit a support case and link it to the thread.

Thanks Again!

Joe

[foggy]

With 7 points retention, the old full with corresponding increments will be deleted only after the number of points in the new chain reaches 7 (this is just how forward incremental backup works). So everything looks expected in your case.

[Lershac]

I have a daily backup with synthetic set for saturdays, forward incrementals, retention set to 7 restore points. Rotating USB external media

I would expect a maximum of 14 restore points on disk, but right now I have 47... am I misunderstanding what "restore points" means?

I have 9 .vbk files and 37 vib files spread proportionately over a couple months date range.

Can I get some clarification on what I am doing wrong? My rotating media is filled up!

Thanks

[Shestakov]

Hello Chuck!
What vesion of Veeam Backup & Replication are you using?
Also please review relevant posts of the topic. Thanks.

[Lershac]

V8-

I use 5 external USB disks rotated on weekdays. m-f.

[Shestakov]

Thanks for the answer, Chuck.
And how many restore points you can see in the restore wizard?

[robg]

Hello, I plan to use RDX media for supplemental backups. My question is how Veeam will treat the data that is already present on a newly inserted tape..

I would like to use the reverse incremental method. Will Veeam:

A: Scan the tape before every job and combine the existing data with the current, to make a proper full backup or
B: Only "remember" last night's full backup, which is on an entirely different media, while the one just inserted has a week old backup.

I am not sure how the two would combine in this situation. While a reverse incremental would run much faster Perhaps it is safer to stick to active full for RDX?

Thanks

[Vitaliy S.]

Hi Rob, please take a look at this link for the answer. Thanks!

[Moebius]

Hello. I'm implementing a Backup Copy job on external USB 3 disks.
The USB disks will be attached, one at a time, to a Synology NAS (which is the primary backup repository) and will be seen by the Veeam server as a SMB share.

I am planning to rotate 3 disks on a weekly basis (say, on Fridays): The Backup Copy job is set to copy daily and to keep 10 restore points (so to avoid the merge phase if for any reasons the "disk swap day" is postponed a day or two). The sequence will be Disk#1 -> Disk#2 -> Disk#3 -> Disk#1 -> etc.

I understand that, when I insert Disk#1 after Disk#3, Veeam will see the chain as broken and will start a new chain, beginning with a full, which is fine. But will the old files already existing on Disk#1 (and two weeks old) be purged automatically in this case?

[foggy]

Lucio, as it is mentioned earlier in this thread, v8 backup copy job is able to continue incremental chain on each of the drives, so existing backups will be subjected to specified retention and merge will still occur accordingly. To avoid merge, you should specify ForceDeleteBackupFiles reg key for existing backup files clean up.

[resruss]

We often use a couple of single-disk NAS devices that are swapped out to perform offsite copies of Veeam backups
In the past, we've run a post-job script that uses robocopy to sync the data

I've seen that Veeam's Backup Copy jobs now support the rotated media, but unsure if this applies to NAS

Am I right in thinking that the rotatable media option can be enabled with a couple of NAS as the target for backup copy if:
- The NAS are configured with same IP address, UNC path, folder structure and credentials (but different hosntames and a label on the unit to identify which by hostname)
- Create the repository as SMB/CIFS
- Enable the Hard drive is backed by rotatable media

What would the backup copy job do during the nightly runs? Create a new VIB based off the changes from the VBK on NAS and the next VIB?

Also just wanting to confirm Veeam version level where this was introduced into the GUI

[veremin]

What would the backup copy job do during the nightly runs?

Full backups.

Create a new VIB based off the changes from the VBK on NAS and the next VIB?

No, that's correct only for Windows-based repositories, not for CIFS ones.

Also just wanting to confirm Veeam version level where this was introduced into the GUI

Version 8.

[motoxrdr21]

What's support like for running a Backup-Copy job weekly to rotating USB3 external drives? I'm working on setting it up in Veeam now and wondering if there are any best practices, known caveats, or whether it's recommended against entirely?

[foggy]

Steve, please review the thread above for known caveats and best practices for rotated drives scenario and feel free to ask if any clarification is still required.

[resruss]

Full backups.
No, that's correct only for Windows-based repositories, not for CIFS ones.
Version 8.

Thanks for those responses.

Are you able to comment on my other query please:
Am I right in thinking that the rotatable media option can be enabled with a couple of NAS as the target for backup copy if:
- The NAS are configured with same IP address, UNC path, folder structure and credentials (but different hosntames and a label on the unit to identify which by hostname)
- Create the repository as SMB/CIFS
- Enable the Hard drive is backed by rotatable media

[foggy]

You can enable rotated drives scenario with NAS boxes, however the full path to the folder where backups are located should be exactly the same (as specified while creating repository). So use IP address to add the repository. Also note that with CIFS-type repository, new full backup will be created each time the new box is connected.

[mashton]

Hi,

Setup:
Three VM Host
One Veeam server located into one of these Host (Virtual)
7 Tb of local disk dedicated for Veeam backup repository usage
600Go of data to backup.
We also have a pool of 16 external usb 3 drive of 2Tb each

example:
(week) HDDPortable-Friday1, HDDPortable-Friday2, HDDPortable-Friday3, HDDPortable-Friday4.
(Month) HDDPortable-Jan, HDDPortable-Feb, ...


1- We do backups of all VMs into the local repository. (Multiple backup jobs for each types of OS. (This is working)

2- We would like to copy those backups into a USB Drive at each friday. (We only need the actual week backups on this USB drive. It will be placed into a safe after that).

It's like a backup to tape, but it's for USB Drive.

Can somebody knows how I can configured this !

Thanks a lot !

Marc


I'm trying to make a job that will do backups of a week into one external USB drive (rotation of USB drive keeped in a safe)

[foggy]

Marc, I'm merging you into the thread discussing rotated drives scenario so that you be informed about all caveats of the current implementation.

Basically, you would need to create a backup copy job with a weekly copy interval and configure it to monitor your source backup jobs for new restore points availability.

[souperstar]

Basically, I would like to know if v8 would now allow me to use a Backup Copy Job for daily rotated drives. Or, if you have another suggestion I'm all ears.

The way it's setup now:
- The backups exist on storage local to the backup server, then they could get copied to daily removable drives using a Backup Copy Job, correct? I'm hoping to fix my "copy TB's of .vbk data to disk every day" issue.
- In v7, I was unable to have Backup Copy Jobs re-use an existing .vbk already on the disks. Can the job do this now with a week old .vbk that's not in the Veeam Backup retention records?

In the v8 "What's New" I noticed this:
Backup Copy suspend
: A Backup Copy job is now suspended when source backup files are locked by
another job or process, such as a restore or full backup file transformation, instead of just failing.
--Good, that fixes the gobs of errors I was getting (hopefully).

Backup mapping
: In addition to full backup, you can now map Backup Copy jobs to any existing
backup chain, as long as it does not contain reversed incremental (VRB) files. This dramatically simplifies
Backup Copy seeding, and makes it possible to repoint Backup Copy jobs after moving backup files to
another storage target.
--Doesn't sound like what I'm looking for, but perhaps I can script it with powershell or something so that I can import an existing .vbk job on the removable disk and have the Backup Copy run afterward?

Simply put, I'm trying to find the most efficient way of copying a full backup file (I favor Rev Inc jobs, but could be persuaded to change) to daily rotated disks that I won't see but one day per week. I'm hoping the Backup Copy Jobs can get me there, as I hear they only need to copy the changes. This would save me significant time over robocopy.