How does Protection Manager handles a back-up retention period > max. (= 255) amount snapshots ?

[dammers]

How does Protection Manager handles a back-up retention period > max. (= 255) amount snapshots ?

So, how do we handle a daily back-up with a retention period of a year ?

[adaikkap]

Hi Robin,

Currently this is not handled.Once the number of snaphots on the

volume has reached the limit of 255, the further protection jobs fails

with the job details for snapshot create giving the following error.

"No more snapshots available"

There are plans to handle this in next release of protection manager.

Regards,

[dammers]

By next release, do you mean v 3.8 ?

[adaikkap]

Not 3.8, but release next to it.

[adaikkap]

Hi Robin,

Not 3.8, but release next to it.

Regards,

adai

[dammers]

Hi Adai,

How can we work aroud this ? (till the appropriate version is released)

Should we create different data sets with different remote back-up schedules,

e.g. one schedule on mo, tue, wed and the other scedule on thu, fri, sat

Then I'll probably need two different secundary volumes, each having a baseline copy + 156 (=3*365/7) incremental Snapshot copies.

Maybe you've got some other suggestions.

Rgrds,

Robin

[adaikkap]

what you mentioned is one way.

i am trying couple of other things as workaround for your issue.

Will let you know once i do some hands-on and share the test results.

regards

adai

[adaikkap]

Hi Robin,

What you said is one way of doing it.Other options are,

Option 1:

Would be to force the volume to migrate after 250/255 snapshots by tweaking the volume full threshold ,there by all backup relations are intact.

But the disadvantage of this would be it would require a base line transfer of the source volume from primary filer, which is a costly operation.

Option 2:

Would be to use flexclone, there by common data is shared no need of double space or atleast the extra space required for the two baseline data.

But this will require fixing the backup relations which we need to try and test before giving this as a solution.

Option 2 would be the superior options compared to option 1 and what you suggested.

Regards,

adai