Direkt zum Inhalt

Backup format 12...not a single file wirth tibx but several with 0001, 0002...

Thread needs solution

Hey there,
I uses 11.7 vor over a year and then about 3 weeks ago I switched to 12.5.

Since these 3 weeks I have more problems and isses with Acronis than in the 19 months (!) before.

My actual problem:
I´m currently testing the storage location "network folder" as the Storage Node makes more probles than it has advantages.

So I changed the storage location in the plan to use a network folder.
The plan has a GFS scheme so I ran manually 6 different runs to check what happens.

My assumption:
- One tibx file and all backups in it.

 

Reality:
- 5 tibx files, the last four are numbered with -0001 to -0004

Huh?

The backup section in the management console shows 6 backups:
- Full backup at 20:35
- Differential backup at 20:40
- Differential backup at 20:46
- Differential backup at 21:58
- Full backup at 22:30
- Incremental backup at 00:30 the next day.

 

The 5 backup files: are dated:
- *.tibx at 20:40
- *-0001.tibx at 20:46
- *-0002.tibx at 21:58
- *-0003.tibx at 22:30
- *-0004.tibx 00:30 the next day.

 

Any ideas why ths strange behaviour is present and how can I solve it ?!?

 

 

Another question:
Why is the archive file that is generated by a storage node located in a subfolder structure and the archive file crated by the agent is directly in the specified stroage location?

 

Regards

Sven

 

0 Users found this helpful
frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Beiträge: 22
Kommentare: 3800

Hi Sven,

When you set up the GFS scheme, the manual runs of the backup task will produce different .tibx files - this is a by design behavior due to GFS scheme specific (since each full backup has to be in a separate physical .tibx file). There is also a difference depending on which type of backup you perform manually (drop-down menu for "Run now" button) and it's unclear which option you selected there in your manual tests (full/incremental/differential). The default behavior for "Run now" button on the backup plan is to produce incremental backup which will be saved into the same .tibx file as the previous backup. If you choose "differential" option in the drop down menu, then it will produce a separate .tibx file and the further incrementals will be saving data into this new .tibx file (000X.tibx).

Here's a screen shot example from my environment with mapping between recovery points and their physical location inside .tibx files - this is all caused by specifics of GFS scheme where separate .tibx files are required for full and differential backups in order to properly support such scheme:

BackupsMapping.PNG

If you want to have a single .tibx file then you need to use "Always incremental" (default) backup scheme. What is the problem with the GFS scheme and keeping backups in multiple files? Which scenario are you trying to implement that you need a single file and "Always incremental" scheme does not fit it?

The files/folders structure in a location managed by Acronis Storage Node (ASN) is specific since the archive format and overal data structure in ASN location is different from backup to plain network shares.

Thank you.

Hello Vasily,

thanks for the answer.
You see the type of backup that I run (full/diff/inc) in my first post:

*********
The backup section in the management console shows 6 backups:
- Full backup at 20:35
- Differential backup at 20:40
- Differential backup at 20:46
- Differential backup at 21:58
- Full backup at 22:30
- Incremental backup at 00:30 the next day.
**********

The presence of the several files can be explained by the things you describe:
Each Full and each Differential backup creates a new file.

So this is okay, but why is the behaviour of the storage node different:
The change I made to the backup plan was only to change the storage location from a location managed by a storage node to a network folder.

But at the "storage node managed" location there is only 1 single tibx file.
My retention rules are "full at each 1st of the month, differential on each saturday and incrementel each day."
So due to my retention rules (tha plan was firts run in the beginning of january) In this single file one full backup (the first in the month), 3 differential ones (on each saturday) and the last 7 incrementals are present.

It seems as if a differential backup does NOT create a separate file if a differential run is executed by the plan.

is this correct?

 

Regards

Sven

 

frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Beiträge: 22
Kommentare: 3800

Hi Sven,

The main difference between Acronis Storage Node (ASN) and plain share location is that the backups retention is performed differently when GFS scheme is used - in case of plain network share the entire backup chains must be deleted when they become subject to deletion (e.g. when they expire): full+diff/incr. dependent on this full backup. In case of ASN location the backups retention is handled by deleting the outdated data inside the archive, so it's not maintained by deleting the backup files, but rather by marking expired data as "deleted" inside the same file. It's just different implementation of retention rules handling. This difference comes from the fact that that the backup files on Acronis Storage Node are not designed to be manipulated manually, while for plain backup locations, there were adjustments made to support such manipulations specifically for GFS scheme when new backup format is used (hence the creation of separate backup files for full/diff backups).

Still can you please clarify the actual scenario you're trying to achieve and why there are concerns about the internal archives files structure? E.g. how are you going to rely on these specifics in your environment?

Thank you.

Hi Vasily,
okay...so the implementation of archive files is different between ASN managed locations and plain network drives right.
 

There are no concerns but only the quetion why the behaviour is different if ASN is used or not.
No problem, I can live with several files - no matter.

Is there a way to export the ASN managed archive and then continue the backup chain on a plain network share?

Could you please tell me where I find the command line reference for the latest 12.5 build ? The ony located on the servers seems to be a bit outdated as not all commands are documented.

Another question:

Is there a possibility to define a subfolder where archives are stored or are the backups always stored flat in the defined network share?
Are there any known issues if I place several archives in one share?

 

Regards

Sven

 

frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Beiträge: 22
Kommentare: 3800

Hi Sven,

>> Is there a way to export the ASN managed archive and then continue the backup chain on a plain network share?

Yes, it is possible - you can do it either via GUI by "Exporting" the recovery point:

ExportBack.PNG

OR by creating a separate backup replication plan (from Plans->Backup Replication) where the source location would be the managed vault and the target is a plain network share.

OR (it's less convenient though, since it will require to perform backup of the machine at least one more time) by enabling 2nd destination option while configuring the backup plan for a machine, where the 1st destination was managed vault, while 2nd destination is a network share.

To continue backup to the replicated archive you'll need to adjust the "Backup Options->Backup file name" in the backup plan to map the archive to the copied one:

Filename.PNG

>> Could you please tell me where I find the command line reference for the latest 12.5 build ? The ony located on the servers seems to be a bit outdated as not all commands are documented.

The command line documentation is available here: https://www.acronis.com/en-us/support/documentation/AcronisBackup_12.5_Command_Line_Reference/#37072.html . If you've already checked it and didn't find required information, please let me know which option you were looking for?

>> Is there a possibility to define a subfolder where archives are stored or are the backups always stored flat in the defined network share?
Are there any known issues if I place several archives in one share?

When you create a location on some folder (local or network), all backup archive files will be generated in this folder - there wil be no sub-folders in there. Storing multiple archives within the same folder is the actual recommended way, so there should be no problems with it (no known issues here).

Thank you.

Thanks for the information.

The method by replicating sounds good....
I will give it a try and come back if I have issues.

 

Regards
Sven