VEEAM and PostGres as Configuration DB

[taurus1978]

Hello VEEAM Team,
we have a VEEAM 12.2 Installation with postGres 15.8 running as the configuration Database. We have a big NAS Environment which we backup using VEEAM NAS features.
Additionally we copy the NAS Backups to a Tape Library for offsite storage.
We observe that VEEAM puts heavy load onto the DB and that DB is growing every day/week. The DB runs on a local NVMe device. And the size is around 670GB right now.

Is it possible to limit the rate of growing.? For expample using postGres Maintanance Tools on a regular basis ? Can you give some advice here ?:
Reference:
https://www.postgresql.org/docs/current ... uming.html



Thank you.

[Mildur]

Hi Patrick

Our file backup engine requires disk space for metadata and indexes. If you can share with me some specifics I can do a short check with our internal calculator if your database size is expected or not:
- NAS source data (TB)
- number of files / folders (in millions)
- number of full backups
- days of retention

Is it possible to limit the rate of growing.?

No, all data in the database is required. You can't just delete datasets from the database.

Best,
Fabian

[taurus1978]

Hello Fabian,
thank you for the quick reply. I will look after the requested data.

"No, all data in the databse is required. You cant just delete datasets from the database."
Sorry to say but this argument seems a little bit to easy for me .

I speak aboput "DB maintance" which (in my opinion) is carried out on the DB by default. (psql = vacuum Daemon)
Maybe it is possible to reflect on those maintenance tasks internally and give out some "dos" and "donts" ?

[Mildur]

Hello Patrick

"No, all data in the databse is required. You cant just delete datasets from the database."
Sorry to say but this argument seems a little bit to easy for me .

I speak aboput "DB maintance" which (in my opinion) is carried out on the DB by default. (psql = vacuum Daemon)
Maybe it is possible to reflect on those maintenance tasks internally and give out some "dos" and "donts" ?

Autovacuum should already be enabled on the PostgreSQL server.

I apologize, I thought you meant that you want to clean up data from tables within the configuration database. Removing datasets from the database is not supported. I suggest reaching out to our customer support team to verify if the database is at the expected size or not. If necessary, we can then initiate discussions with our QA team to explore possible optimizations.

Best,
Fabian

[taurus1978]

Hello Fabian
we opended a case with support. Case Nr. is #07443080
Can you please initate a discussion with the QA Team ? We need to investigate this further.
Current issue is that VEEAM Console take too long to open an review the environment.
Thank you.

[Mildur]

Thank you for the case number.
Can you please also let support know about the data I requested in my first comment?
We need to verify if the database size is expected or not before I can reach out to QA.

Best,
Fabian

[taurus1978]

Hello Fabian,
can you give advice on how to collect this information ?. We dont have access to cusotmers Fie Environment.
Thx.

[Mildur]

Hi Patrick

- NAS source data (TB)
- number of files / folders (in millions)

This information should be retrieved directly from the source system if possible.
But you can also check your job statistics from a full backup run. There is a line in the session details which tells you how many files and folders were protected by this job.

- number of full backups
- days of retention

This are settings in the backup job.
How many full backups do you keep? How many daily restore points?

Our configuration database will grow a lot depending on the amount of Files/Folders you backup. That's why I want to verify if 670GB is an expected size. We have it in our help center: https://helpcenter.veeam.com/docs/backu ... ase-server

The database space required for processing 1 000 000 files and folders is 850 MB.

The database will grow high if the Fileserver has millions of files/folders. And the size has to be counted for every new backup chain.

An example from our internal calculator excel:
- 500 million files/folders
- Keep 2 full backups (2 is minimum)
- Keep 60 days of retention
The database size requirements for that single job will be approximately 1TB.

Best,
Fabian

[taurus1978]

Hello Fabian,
i want to give a follow up on this and to provide some data to the environment. We do File backup to Disk and also to Tape
On disk we keep 14 restore points. We do a backup copy to an immutable repository and keep that also for 14d. We also do a daily tape out and keep that for 2 days. Additionally we do a weekly tape out and keep that for 26 weeks.
The biggest file share in the environment has around 40 million folders and 20 million files. (Yes ist big) . The other shares are much smaller than that.We do not have a detailed listing of the shares.