Storeonce + copy jobs

[kte]

How can I create a copy jobs that keep 31 days on a windows server DAS repository ? and store the GFS files on a storeonce device ?
Can I use scale-out for this and is there an option to put the GFS weelky and monthly fulls on the storeonce only , how can this be done?

[foggy]

That's exactly where scale-out repository can help you. Please review this user guide section as a starting point.

[kte]

So do i use performance or locality, because i want my normal backup chain to be togeteher and only the GFS full backups on my deduplication appliance?
not sure this is possible, because my local veeam presales said it is not possible.
can i mix the same repositories for regular copy jobs an use the same repository for scale-out backup copy jobs

[foggy]

Koen, this is not possible, you cannot separate regular backup copy chain from its GFS restore points.

[kte]

so what is the best practice but the full backup copy jobs and GFS on the sroreonce
or create a second copy job with just the GFS , can you create 2 copy jobs from 1 backup job

[foggy]

Yes, you can create two copy jobs from a single backup job. What is the specific reason you want to have an entire regular chain stored on a single extent, separately from GFS points?

[kte]

to be able to do surebackups, fast restores from a normal data store during 31 days from the copy jobs and archiving on a deduplication appliance with GFS only

[foggy]

SureBackup is not available for backup copy jobs anyway and fast restores are also typically performed from the primary repository.

[kte]

we keep 14 days on the production site and 31 days of copies on the DRP site and there the wand to keep weekly and monthly fulls, so gfs on the storeonce should be ideal because it could be we need a restore from the DRP site from the 31 days copies

[foggy]

Understand. Then you can set up two backup copies, one specifically for fast restores from 31 restore point, another specifically for GFS with the shortest regular retention possible (2 restore points).

[mkretzer]

Hello,

why is it not possible to keep the GFS restore points on a different extend in a SO repo? It would be nice to be able to specify that the GFS points go to a specific extend so that we can enable MS dedup there. MS dedup on a repo which is a copy target leads to endless tries to dedup the ever-changing backup file.

The other idea would be to be able to have a incremental GFS - in many VMs data is constant even over a long time.

Markus

[Vitaliy S.]

Markus, thanks for your feedback! In order to achieve your goal today, you can use a separate backup copy job to store files on a different repository (extend).

[mkretzer]

After thinking about this a little bit longer i think i missed one feature: We can set MS dedup to only dedup files older than a few days which should lead to it only "deduping" the GFS backups.

Which now only leaves the size restrictions of MS dedup...

Markus

[Vitaliy S.]

Good point! As far as I know there are lots improvements coming to Hyper-V 2016 in regards to dedupe engine (even though it doesn't help you right now).