bootstrap

[jlh16]

Hi all,

We are running Networker 7.1.1 on a Solaris 8 machine. During a backup the bootstrap was being saved to a tape that was full. It then was saved to a different tape. The savegroup completion report contains the following error:

savegrp: servername:bootstrap bootstrap was never started

I know that it backed up from the daemon.log entry:

06/13/04 08:34:55 nsrd: servername:bootstrap saving to pool 'Weekly' (000000L1)
06/13/04 08:34:56 nsrmmdbd: media db is saving its data. This may take a while.
06/13/04 08:35:14 nsrmmdbd: media db is open for business.
06/13/04 08:35:17 nsrd: media warning: /dev/rmt/2hbn opening: wfm failed: drive status is The end of medium was reached
06/13/04 08:35:17 nsrd: media notice: LTO Ultrium tape 000000L1 on /dev/rmt/2hbn is full
06/13/04 08:35:17 nsrd: media notice: LTO Ultrium tape 000000L1 used 103 GB of 100 GB capacity
06/13/04 08:37:58 nsrd: media info: verification of volume "000000L1", volid 1389103575 succeeded.
06/13/04 08:38:12 nsrd: write completion notice: Writing to volume 000000L1 complete
06/13/04 08:41:13 nsrd: servername:bootstrap done saving to pool 'Weekly' (000001L1) 21 MB

It appears Legato got confused because the bootstarp is listed when viewing the volume "000001L1." Any suggestions on how to make sure it doesn't get "confused" again?

Thanks,
jlh16

 

[605]

My earlier experiences showed that the bootstrap must not necessarily reside completely on a media - as with any other save set it may span on the next media.

But the report looks ok: NW could not write it on volume 000000L1 and does it on volume 000001L1.

The only problem is the "bootstrap was never started" message which obviously is a bug in this situation.

 

[jlh16]

It wouldn't be a problem, except that we run reports using the savegroup completion report. The report does not complete successfully when this "error" appears. This is the second time it has happened in the last 3 weeks. It has only happened in our weekly backup which runs on Sunday's. I will have to poke through the program that produces the report and find out what is stopping it from completing the report.

 

[DavinaTreiber]

Unrelated to your problem, why are you using the hbn device? I notice the tape filled at 103GB, this is basically only native capacity. If you change to the cbn device you will probably get more like 200-250GB per tape, and correspondingly better write speed.

 

[jlh16]

We don't remember why it was originally set to hbn. I remember someone from Legato telling us that if a tape was written on hbn it couldn't be read on cbn. We do use compression on the client side.

I have placed a call to Legato about the bootstrap. They have bumped it up to the engineers as a possible bug.

 

[DavinaTreiber]

Well your "someone from Legato" was wrong. Anyway if you stick to cbn for everything you can't go wrong.

 

[cstorm]

Legato Networker gurus,

Is there a way to configure the server to write its bootstrap files to a storage node with tape drives?

 

[605]

Negative. This is a 'hardcoded' feature. Although some of you might see this as a limitation, it makes sense. Just think about where you want to have the data in case you must recover the NW server.