Tapes not expiring

[chekcol]

I am having an issue with a couple of tapes not expiring. There are all FULL backups on the tapes so shouldn't have dependencies that is keeping them around. The browse and retention perriod is set to 1 month. In this case the save sets were written to tape on Oct 14 and should have expired Nov 14 but not the case.

The version of NW is 7.3.2 and running Alphastor 3.1.

Any help would be great.....thank you

 

[denisfr]

Hello,

have you checked all save sets retention (mminfo -avot -r "client,name,ssretent" vol_name) ?

have you some saves of your networker server on it ?

 

[chekcol]

I will run that command and see what I get. The save sets in question run a full backup everyday, it is the backup server so i understand it will keep data until another backup cycle is run. Would the backup cycle be one day?

Thank you

 

[chekcol]

ok...i ran that command and all the savesets are shcheduled to expire on 11/14 but the tape has not recycled itself

 

[denisfr]

When networker saves its data, it keeps retention according to the longest retention it can find in all definitions of itself.

e.g : you have your server which is client of himself, maybe multiples times, you have a save with 3 months retention, an another with 2 months. If you look at the retention of saves made with 2 months retention, you may notice that retention is 3 months. This problem/feature is in the todolist of EMC (RFE).

hope my poor english will help.

 

[chekcol]

Ok thanks.

That makes sense and the case for this one clinet but I just found another client that has only one defined and browse and retention is 1 month and should have expired on 11/13. I did notice there is a bug in 7.3.2 LGTPA91485 that is fixed in 7.3.3. Just want to make sure before I tell the client.

 

[605]

I am a bit worried about this statement:

" .. all the savesets are shcheduled to expire on 11/14 but the tape has not recycled itself."

It makes me think that you did not get it right.


The fact that a NetWorker media is recyclable does only mean that it MAY be recycled. NW will not recycle/relabel it right away - it will do it only if it needs a tape (from the same pool (default setting)).

The intention is clear: this ensures the longest period for potential recoveries - as long as the media has not been overwritten you may still retrieve your data.

 

[denisfr]

I just found another client that has only one defined and browse and retention is 1 month and should have expired on 11/13"

What I explained is only applicable for the client 'networker server', not for clientA or clientB.

You said you found another client which may have expired. Did you check ssretent or volretent for this client ?

Try 'mminfo -av -r "volretent,state" your_vol_name' to see, as 605 points to it, the state of the volume.

 

[oldnixguy]

605 --- Does this mean, then, that when a volume is "undef" that NW will grab it and mark it recyclable as soon as it needs it (when there are no append, or recyclable currently volumes available)?

 

[605]

I do not understand what you mean with "undef" - such volume status does not exist.

If ALL save sets for a media are RECYCLABLE, then NW will set the media to RECYCLABLE to indicate that it MAY now be overwritten. This is just a DB operation - NW will not do anything to the media at this point in time.

If NW is writing AND if there is no appendable media in this pool, then NW will relabel and use it for writing.

 

[oldnixguy]

state volume written (%) expires read mounts capacity
10000E 223 GB full undef 115GB 121 160 GB

This is the output of "mminfo -q "pool=Nix Pool" -m"

I'm trying to figure out the "undef" reference. I get groups of tapes that cycle through as undef.

 

[605]

This value refers to the mminfo parameter "volretent". AFAIK it is not used any more. The latest retention date of all save sets will be used instead. So you better look for the volume status recyclable.

This, unfortunately, does not help you if you want to know when the volume will become recyclable.

 

[oldnixguy]

Thanks ... I may have been wasting my time chasing that undef value!