Networker 7.3.2 - No free tapes - Urgent Advice required please 2

[woodings]

Please help - currently using Networker 7.3.2 - ADIC I2000 robot with 6 LT03 drives with 200 LT03 tapes - backup retention policy is one month - unfortunatley with the problems of Dynamic Drive sharing (it was recently introduced to our site), we now have no free tapes. I know this beggars belief , what would be the quickest way to free up some tapes - would it be to go back and check the oldest savesets and then relabel tape for reuse, or is there a better and more correct method

 

[605]

You can use manual staging (nsrstage -s ssid -b pool) to migrate save sets to another media. Then relabel the source tapes.

 

[Rif123]

If buying more tapes are out of the question and there is no tapes to rotate into the robot, I would do what you just said.

There is no good way to do it since you will end up and ruin the backup cycle. Hopefully you can keep your SLAs in that case.

So if you take the oldest tapes and recycle them, bear in mind that doing so can and will wreck your backup cycles if you are using incremental backups. Maybe you have different pools so you can select less important tapes or database backups that always runs full.

 

[lotamoka]

We are backing up EMC nas as NDMP backup in legato. Is it possible to backup NETAPP nas in the same networker backup server.

 

[605]

Yes - the best is probably to use the new DSA (Direct Service Agent). It is included in NW since NW 7.2 and does not cost a penny. And all other limitations which you had for NDMP backups are gone.

 

[woodings]

An additonal problem is that the LT03 tapes are being marked as full, when they are clearly not - for example when looking at the individual volumes, as little as 5GB of data has been written to a tape. The tape is then marked as full. I cannot understand this - the tapes have varying amounts of data written to them, some are full others are not. I cannot see a particular pattern which would cause this to happen.
Our configuration is ADIC Scalar I2000 robot with 6 LT03 drives, 200 tapes utilising dynamic drive sharing - we have a Networker Unix server which is the storage node and two Windows 200 dedicated storage nodes - all devices can be seen by each node - it was suggested that we check hardware compression was enabled on windows drives at the windows level - it is!. Also within Networker we were adivsed to changed the default block size on each drive to 128kb, but leave the 6 unix drives at handler default.
This does not appear to have made any difference - tapes are still being marked as full when not, and tapes are not being expired when they reach their expiry date.

 

[605]

When NW can not recover from a write error, it marks tapes as full prematurely. It is not easy to determine whether this is a drive or a media problem. Maybe the log will show you some sense code and you have a hardware manual which could explain the problem.