Several weeks writing a weekly work tape?

[infrasebraego]

I have backup jobs for daily, weekly and monthly tapes. LTO8 model. There is a lot of free space left on each tape, and I would like each tape to be able to store several days, weeks and months. For example, the weekly tape can contain several weeks of recordings if it fits.
Is this possible?
I have already researched and have not received an answer.

[david.domask]

Hi infrasebraego, welcome to the forums.

Can you describe a bit more on your tape-out strategy?

1. Are you using Tape GFS or just regular Backup to Tape? The behaviors are a little different, and it sounds like Backup to Tape is what you're hoping for, but if you want archival points (GFS) on tape, you will need to use Tape GFS. You mention Daily, Weekly, and Monthly, but just want to confirm if it's a Tape GFS job or multiple Backup to Tape jobs. (You can tell if the media pool you're targeting is a GFS one or a "Simple" media pool)
2. I think what you want is the Append option for GFS media pools if you're using Tape GFS; Simple Media pools have a similar option with the media set configuration "Do not create, always continue"
3. I think media sets are probably what may be unclear here for you. See our FAQ on tape here for a deeper explanation on how media sets work.

Thanks!

[infrasebraego]

The current strategy implemented by the supplier is as follows:

We have a common disk job that runs daily, with GFS configured, and performs the full synthetic on Saturdays.

Next, we have backups to tape jobs, where we select the jobs to be copied to specific tape pools. We have specifics jobs "backup to tape" for daily, weekly, monthly and yearly tapes, and we also have specific pools for daily, weekly, monthly and yearly.

The "Do not create, always continue using current media set" option was chosen, but the tapes are taking up too much free space.

I don't know if it's the best strategy, but we're new to Veeam, and it was the supplier's suggestion.

I noticed that the data written to the weekly jobs is about 5TB, half of what the LTO8 tapes support, and we're finding that there's a certain waste of tapes.

I don't know if I've provided enough information, but I'm available for more details.

Thanks in advance.

[infrasebraego]

I hope this print helps.
The first image is of the weekly tape job, where we selected the disk job as the source.
The second image is the tape pool chosen.
The third image is to show the configured pools.
Prints in Portuguese because I'm in Brazil.

link for prints:
https://ibb.co/LSqY4qB

[david.domask]

Thanks for the details infrasebraego; looking at the screenshots and your settings, it sounds like all is configured as expected; for the first tape in the media set (SE0001L8), looks like it was filled pretty much to the end of tape (only 4.4 GiB remaining, which is normal as usually the End of Media marker on tapes is set somewhere near the end of the tape with < 10 GiB typically left).

The next tape SE0002L8 looks like it's the current tape in the media set, and on the next run, presuming the tape is available in the library, Veeam will try to use it and will continue to use the tape until either it is filled or the media set gets closed. (Note the conditions when a media set may close automatically outside of the scheduled time)

In short, the next run should use SE0002L8 and continue writing where it previously left off -- are you seeing different behavior in your tape jobs currently? Are the tapes being exported by chance either automatically or manually in between job runs? That would close the previous media set as noted in the User Guide link above.

[infrasebraego]

are you seeing different behavior in your tape jobs currently?

Yes. The daily pool tapes, in particular, are taking up a lot of space.
Even with the same settings as the weekly pool.

Maybe after the images expire it will work as per my expectations?

Image Link:
https://i.imgur.com/HSgG6ug.png
https://imgur.com/a/T2aQRgR

[david.domask]

In that case, if you're seeing different behavior, please open a Support Case and let Veeam Support review the logs for the affected jobs.

My guess is that it is related to media sets closing (possibly due to being ejected and not returned for next run), but Support will be able to explain the behavior in detail after a log review.

Please share your case number here once the case is opened. Thanks!

[infrasebraego]

Thank you very much david.domask, you really helped me a lot with your comments. I am immensely grateful. I found the "GFS Media Pool" option fantastic, which I did not know about, and I am even thinking about migrating to this model, I just need to analyze it first.
I will wait for the images to expire and see how they behave. If this issue is not resolved, I will open a ticket with Veeam. In both cases, I promise to return here and report the results of my experience.

[infrasebraego]

I'm just coming back to report the results.
After the first expiration cycle, the jobs started to make the most of the tape space as expected.
I don't remember if "Create new media set...." was selected at the beginning and caused the strange behavior, but now everything is working fine.
Thank you.

[david.domask]

Hi infrasebraego,

Thank you for the follow up and glad to hear it's now working in alignment with your expectations, and glad to help!