mix problem of clone pool and stage

Its a bit unclear from your description what is happening here. If you backup some data it must go to a backup pool; if you clone this data it must be cloned to a clone pool. However, if you stage the data whether from the original or clone, it does not matter what type of pool you stage the data to.

TO answer your question, you can have more than one instance of a saveset in a pool, but you cannot have more than one instance on the same piece of media. So it is not the cause of this error.

Since you are implying you have several pieces of media in the pool you are staging to then I would not expect it to be an issue with an instance of the saveset being on that tape (since it is unlikely the saveset has been copied multiple times to multiple tapes). I would suggest you verify you have appendable tapes assigned to the pool you are staging to; you did not indicate whether or not manually labelling/mounting the tapes caused the stage to complete so it would be useful to know this. There may be issues with the media database so I suggest you run nsrim -X from the command line to run a check on your media database.

How do you stage? Do you use script or NW? Do you have more, less or equal number of tape devices as opposed to number of disk devices? If you have 5 disk devices and 4 tape devices and staging kicks in for all 5 this is normal message and it will be there until one disk device is done with staging and can continue with 5th one.

Its a bit unclear from your description what is happening here. If you backup some data it must go to a backup pool; if you clone this data it must be cloned to a clone pool. However, if you stage the data whether from the original or clone, it does not matter what type of pool you stage the data to.

I backup the data to Adv_File media first and then stage the data to clone pool if the capacity over the watermark. I think this destination pool type must be a clone pool, right?

And I also enable auto clone function of schedule backup, all data will be cloned to the same pool.

The loop error message as below:
4/13/06 04:15:26 nsrd: media request alert: Waiting for 1 writable volumes to backup pool 'Local Clone' disk or tape on legato.bop.com.tw
04/13/06 04:15:49 nsrd: [Jukebox `ADIC@3.2.1', operation # 99]. Finished with status: succeeded
04/13/06 04:16:14 nsrd: \\.\Tape0 Verify label operation in progress
04/13/06 04:16:15 nsrd: \\.\Tape0 Mount operation in progress
04/13/06 04:17:28 nsrd: media notice: Volume `000018' ineligible for this operation; Need a different volume from pool `Local Clone'
04/13/06 04:17:28 nsrd: media request alert: Waiting for 1 writable volumes to backup pool 'Local Clone' disk or tape on legato.bop.com.tw
04/13/06 04:17:32 nsrd: [Jukebox `ADIC@3.2.1', operation # 100]. Finished with status: succeeded
04/13/06 04:19:30 nsrd: Operation 101 started : Load volume `000018'.

I can't understand why it always to load the Volume 000018 and said it showed it will need more writable volumes.

I think you have correctly figured out that the tape it is trying to write to already has a copy of that saveset on it.

Can you confirm that you have other appendable volumes available from the same pool? If you manually unmount volume 000018 a load another appendable tape from that pool do we see things kick off or does it still say waiting for a tape?

Have you also tried using a different pool for either your clones or your staged data? If so did things work without problems?

How do you stage? Do you use script or NW? Do you have more, less or equal number of tape devices as opposed to number of disk devices? If you have 5 disk devices and 4 tape devices and staging kicks in for all 5 this is normal message and it will be there until one disk device is done with staging and can continue with 5th one

I choose the NW stage to use.
Could you explain what's different between script stage and NW stage?

Thanks.

Can you confirm that you have other appendable volumes available from the same pool? If you manually unmount volume 000018 a load another appendable tape from that pool do we see things kick off or does it still say waiting for a tape?

The Scalar24 tape library havs some expired volumes and I ever tried to munually label one or two volumes for use. But the message still appeared even if I restart the process or the backup server.


Have you also tried using a different pool for either your clones or your staged data? If so did things work without problems?

Does this happen during staging or cloning (or both processes are active)? So far I always used to see different pools for cloning and staging so I would agree that is a good starting point for further testing.

Not too much - manual one gives you more control over when, what and where.

I am using a different pool to test and will post the result.

My guess is that you have got appendable volumes here but not ones in the correct pool or possibly not tapes currently in the tape library. Networker will not relabel tapes into a different pool unless this is configured in the pool resources.

Given that this is small library, when this happens again please send us following:
- nsrjb -v output
- mminfo -av -q location="library name" -r volume,pool,state
- part of daemon.log starting before issue started (without any additional editing)
- output of nsrjb -h

I reviewed the daemon.log and found following error:

04/14/06 14:30:45 nsrd: \\.\Tape1 is now write protected
04/14/06 14:31:15 nsrd: \\.\Tape1 Verify label operation in progress
04/14/06 14:31:19 nsrd: \\.\Tape1 Mount operation in progress
04/14/06 14:31:19 nsrd: media event cleared: confirmed mount of 000005 on \\.\Tape1
04/14/06 14:31:20 nsrd: media info: can not read record 0 of file 2 on LTO Ultrium-3 tape 000005
04/14/06 14:31:20 nsrmmd #2: inactive ssid
04/14/06 14:31:20 nsrmmd #2: inactive ssid
04/14/06 14:31:20 nsrd: media emergency: inactive ssid
04/14/06 14:31:20 nsrmmd #2: inactive ssid
04/14/06 14:31:20 nsrd: cloning session:3 of 17 save set(s) reading from 000005 116 GB of 1932 GB
04/14/06 14:31:20 nsrmmd #2: inactive ssid
04/14/06 14:31:20 nsrmmd #2: inactive ssid
04/14/06 14:31:20 nsrd: media info event: LTO Ultrium-3 tape 000011 will be needed
04/14/06 14:31:20 nsrd: cloning session:1 of 17 save set(s) starting read of 1932 GB
04/14/06 14:31:20 nsrd: media event cleared: LTO Ultrium-3 tape 000004 not needed
04/14/06 14:31:20 nsrd: media event cleared: LTO Ultrium-3 tape 000012 not needed
04/14/06 14:31:20 nsrd: media waiting event: waiting for LTO Ultrium-3 tape 000011 on legato.bop.com.tw
04/14/06 14:31:24 nsrd: [Jukebox `ADIC@3.2.1', operation # 37]. Finished with status: succeeded
04/14/06 14:31:30 nsrd: Operation 38 started : Load volume `000011'.
.
04/14/06 14:31:32 nsrd: \\.\Tape1 Eject operation in progress
04/14/06 14:32:54 nsrd: \\.\Tape1 Verify label operation in progress
04/14/06 14:32:56 nsrd: \\.\Tape1 Mount operation in progress
04/14/06 14:33:00 nsrd: media event cleared: confirmed mount of 000011 on \\.\Tape1
04/14/06 14:33:00 nsrd: legato.bop.com.tw:cloning session saving to pool 'Remote Clone' (000014)
04/14/06 14:33:00 nsrd: cloning session:1 of 17 save set(s) reading from 000011 2435 KB of 1932 GB
04/14/06 14:33:00 ansrd: ansrd_clone FAILED: errnum is -7 and errstr is can not read record 0 of file 108 on LTO Ultrium-3 tape 000004
04/14/06 14:33:00 ansrd: failed to execute MODE_CLONE
04/14/06 14:33:00 nsrd: legato.bop.com.tw:cloning session done saving to pool 'Remote Clone' (000014)
04/14/06 14:33:00 nsrd: cloning session:save sets done reading 1878 GB
pools supported: Remote Clone;
04/14/06 14:33:02 nsrd: write completion notice: Writing to volume 000014 complete
04/14/06 14:33:04 nsrd: [Jukebox `ADIC@3.2.1', operation # 38]. Finished with status: succeeded
04/14/06 14:36:13 nsrd: Operation 39 started : Unload jukebox device `\\.\Tape0'.
.
04/14/06 14:36:13 nsrd: Operation 40 started : Unload jukebox device `\\.\Tape1'.
.

one or two volume has reported the inactive ssid, how to cause this?

When I run the clone script, the system showed that.
my sccipt content as below:
nsrclone -S -e now -t yesterday -g "Default" -b "Remote Clone"

If possible I run the schedule backup using two tape drive simultaneity and run nsrclone may casue the problem.
because I just have only two tape drive?

Thanks,
Dennis

Hi Dennis,

What you have sent now has nothing to do with your first problem - jumping from one thing to another won't solve anything (however from what you have sent I can't see anythin related to backup as obstructing dark force killing your clones).

I believe this is a good moment to open a support call with the SE or EMC (depending on your contract) and use that support. Did you have this problem before 7.3? Btw, I'm also not big fun of the nsrclone command you use (-e switch). Second, given that you have only 2 drives, why don't you clone from disk->tape? What you have sent as a log shows tape->tape operation...

I think it would be useful to see output of the command:

mminfo -q location="ADIC@3.2.1" -r "volume,pool,state"

so that we can see what tapes are in your library and in what pool.

I would suggest to troubleshoot this further identify one saveset to clone and run some tests with that outside of normal backup/clone activity. This will allow you to identify what is going on in the logs a lot clearer (the excerpt you sent from daemon.log seems to have several activities taking place at the same time which makes it harder to indentify what is contributing to our problem). For the test ensure you have the tape with the saveset loaded and an appendable tape in the correct pool also available in the library (you could use a test pool and label one tape into it so that you can keep your testing seperate from your production clones).