Unsolved

This post is more than 5 years old

253 Posts

3634

December 13th, 2007 02:00

Cannot allocate the 1 required device(s)

Hi all,

We are having a problem with a new installation. The new environment has 1 Backup Server and 10 DSN (all of them Windows 2003 Server SP2) which share by FC a jukebox DELL ML6000 and 2 drives IBM LTO-3. The issue is based on following:

There are two volumes mounted on DSN Tapes (or Backup Server), without tasks, when a backup of another DSN is executed with a request of different pool to mounted volumes (if the request needs the same pool to mounted volumes, 7.3.3 does a logical change between path and doesn't unload physically so the backup starts rightly), NetWorker shows the message error:

nsrd: [Jukebox `i500', operation # 6]. Automatically terminating operation `OP_LOAD', instance 6, on jukebox `i500'. Cannot allocate the 1 required device(s).
nsrd: [Jukebox `i500', operation # 6]. Finished with status: failed

and the backup is queued.

Only when NetWorker has to unmount automatically the volumes it's when the error appears. If the Inactivity Timeout is reached or if we run a nsrjb to unmount the volumes (for instance nsrjb -HENvvv), the tapes are unmounted and the backups starts.

Next we write you the done test:

1) Upgrade version 7.4. The same error.

2) Install 7.2.2. The message was the following:
"All of the devices are in use by nsrmmd"

We checked how periodically a nsrjb appeared in the taskmanager to unmount the tapes, but it disappeared quickly. I suppose that this command checked the nsrmmd's state and got an answer telling it they are busy but that is wrong.

3) The RSM Service is stopped in all DSN

4) FW in all components is the newest.

5) No errors appear in Microsoft Event Log.


Thanks in advance.


*We can ensure the pools are rightly configured.

455 Posts

December 13th, 2007 02:00

Check the properties of your devices and make sure DSN is selected.
Check you Pool configuration, ensure all you storage node / devices are properly allocated.
Check you groups and ensure they are allocated to the correct pool.
You may have to reconfigure you library.

455 Posts

December 13th, 2007 02:00

Check the properties of your devices and make sure DSN is selected.
Check you Pool configuration, ensure all you storage node / devices are properly allocated.
Check you groups and ensure they are allocated to the correct pool.
You may have to reconfigure you library as well.

455 Posts

December 13th, 2007 03:00

Check your "Timers" on you library properties, try changing the "idle" time to 1. Also do a Reset on you library.

253 Posts

December 13th, 2007 03:00

All done previosly.

253 Posts

December 13th, 2007 03:00

RSM was already disabled on all hosts configured as SN/DSN

455 Posts

December 13th, 2007 03:00

The problem could possible be with the way your FC switch is configured. I am not sure if you have "Persistent Binding" on your switch, in which case you SCSI Id may have been swapped if you host is restarted.

I suggest you create a case with support and forward them a copy of your Daemon log.

455 Posts

December 13th, 2007 03:00

Try disable "Removable Storage" service on your all you clients

455 Posts

December 13th, 2007 03:00

Try disable RSM on all hosts configured as SN/DSN.

253 Posts

December 13th, 2007 03:00

Idle Device Timeout to 1 is a workaround that at the moment works fine but we want to know what is the source of this error.

253 Posts

December 13th, 2007 03:00

3) The RSM Service is stopped in all DSN (stopped = disabled, sorry)

Are you sure that we must disable the RSM on all clients? I think only on DSN is enough.

null

253 Posts

December 13th, 2007 04:00

I think Persisting binding uses to configure on the host's HBA, I never had listened that it may be configured in the FC switch.

I opened case with support 2 days ago including all necessary information but we haven't got any response yet...

Thanks for your collaboration.

253 Posts

December 19th, 2007 08:00

Has anybody had the same problem? We are still waiting for an answer from Support... and we'd appreciate very much your collaboration

253 Posts

December 28th, 2007 01:00

Yesterday, We changed the jukebox for a new one but the issue is the same. Could be the problem's source that our customer has installed EMC Powerpath on both HBA's?

1 Message

February 4th, 2008 01:00

FYI I have open a case two months ago because of the following :
"
01/31/08 02:14:58 nsrd: [Jukebox `SL500', operation # 477]. Automatically terminating operation `OP_LOAD', instance 477, on jukebox `SL500'. Cannot allocate the 1 required device(s).
"

I have answer yet ... Even if I am aware it's harmless, I would like to know the root cause.

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 4th, 2008 14:00

Generally that happens when load or unload operation can't be addressed by devices available for pool required. Such operation will be retried after.
No Events found!

Top