Unable to clone manually using NMC or nsrclone

You should at least specify the destination pool. If not, NW will determine it automatically.

As you usually have assigned devices to a pool, NW would then use the wrong devices and these can be attached at the wrong (NDMP) storage node.

Hi,

As per error issue with drive (drive selection)  or Media pool.

Check media pool created for it and allocate correct drive for it.

Moin

I do choose a destination pool on the NMC, and on the command line, here's what happens:

nsrclone -s bakserver -b "TEST CLONE" -S 3142953149

80470:nsrclone: Following volumes are needed for cloning

80471:nsrclone:         CAN078 (Regular)

39077:nsrclone: error, no matching devices on `netappfiler'

39077:nsrclone: error, Cannot open nsrclone session with bakserver.  Error is 'No source device is available for the cloning operation.'

79625:nsrclone: Failed to clone any Regular save sets

39077:nsrclone: error, Cannot open nsrclone session with bakserver.  Error is 'No source device is available for the cloning operation.'

39077:nsrclone: error, Some of the clones failed

Yes, but why would the netappfiler want to mount the source volume in order to do a recover clone of a regular saveset?

How do you correct this configuration issue? Which NetWorker resources should be corrected?

Thank you.

What is says is that storage node netappfiler can't mount your source volume.  It also says that source volume is regular type so this was either DSA backup or regular one (in which case filer is probably not where you want to clone of course).  This is typical configuration issue.

It all depends how this is set (recover storage node in storage node client resource) and from where it is run (read hostname). nsrclone has options for this to be set; for example -J will set recover storage node (read) and -d target storage node (write).

The recover storage node in storage node client (NetWorker server) resource is empty check 10-10-2013 16-32-13.png

The manual clone operation is run from a client computer which has the NetWorker Management Console installed.

On the command line with nsrclone, if the -J and -d are specified, it does work.

# nsrclone -s bakserver -J bakserver -d bakserver -b "TEST CLONE" -S 3142953149

80470:nsrclone: Following volumes are needed for cloning

80471:nsrclone:         CAN078 (Regular)

Oct 11 12:18:46 bakserver root: [ID 702911 local0.alert] NetWorker media: (waiting) Waiting for 1 writable volume(s) to backup pool 'TEST CLONE' tape(s) on bakserver

There you go... so now trick is that you set this in NMC too and you are done (and yes, that includes recover storage node too - otherwise some defaults are used which due to device/architecture restrictions won't work for you).

But how do you exactly do that?

If I do a verbose command line nsrclone, this is what I get:

# nsrclone -vvvvvvvv -S 3243714141

Job 'nsrclone of group ' has jobid 7872650.

Obtaining media database information on server bakserver

generate_snodes_from_ss:

parms val: 3243714141

add_to_snodes:

loc_volid:

  volname: CAN026

  using v_location: LIB1

  volume_location: netappfiler.

  hostname netappfiler added

  volid 3897971355 added

vallist_add_ssid:

  ssid: f069a42d-00000006-c157265d-5257265d-01400014-4fc479a8

  cloneid: 1381442967

80470:nsrclone: Following volumes are needed for cloning

80471:nsrclone:         CAN026 (Regular)

79632:nsrclone: Cloning the following Regular save sets (ssid/cloneid):

        3243714141

5874:nsrclone: Automatically copying save sets(s) to other volume(s)

79634:nsrclone:

Starting Regular cloning operation...

6217:nsrclone:   ...from storage node: netappfiler

39077:nsrclone: error, no matching devices on `netappfiler'

39077:nsrclone: error, Cannot open nsrclone session with bakserver.  Error is 'No source device is available for the cloning operation.'

clone_ids before:

        1381442967

clone_ids after:

        1381442967

cloneid in new list = 1381442967

cloneid 1381442967 in old list matched

generate_clone_result(): 0 savesets successfully cloned.

generate_clone_result(): failed to clone 1 savesets.

79625:nsrclone: Failed to clone any Regular save sets

39077:nsrclone: error, Cannot open nsrclone session with bakserver.  Error is 'No source device is available for the cloning operation.'

39077:nsrclone: error, Some of the clones failed

snl_src_snode: netappfiler

loc_volid:

Any clues as to why does this volume belong to a NDMP storage node even though the volume has no NDMP savesets? And how to correct this behavior?

Why don't you read what has been said already?  It depends on read hostname field which is used if no config is specified and if you wish to call it from CLI, you must set parameters as explained.  If you want to schedule it, make sure configuration is done correctly too (set storage node fields for read for storage node client and if using scheduled cloning, you can set some parameters there too).

It seems that the storage node to read data from is generated automatically accordingly to some criteria.

If I clear the volume location field using mmlocation and then re-run the clone operation, it will force it to read the "recover storage node" field of the client from which the saveset is to be cloned.

Then you have to do a nsrjb -I of the tape volume in order to set the location field to the jukebox and then the tape is loaded into the correct storage node and the clone operation takes place.

What I still have to find out is why the storage node generation defaults to the NDMP storage node.

I've read the chapter of the NetWorker Cloning Integration Guide and it states that the read source storage node is chosen based on the following criteria:

The list is based on the storage nodes that meet the following criteria:

a. Storage nodes are listed in the Recover Storage Nodes attribute of the Client

resource of the NetWorker server. If this attribute is empty, the Storage Nodes

attribute of the NetWorker server is used.

b. If the requested volume is in a media library, the storage nodes on which the

volume can be mounted is determined in the following manner:

– The storage node listed in the Read Hostname attribute for the library

resource is used.

– If the Read Hostname attribute for the library resource is not set, then all of

the storage nodes on which any device in the library is configured are

added to the list of Eligible Storage Nodes.

It seems that it is this last criteria that is being applied. Setting the "Read hostname" on the library containing the source tape doesn't fully solve the problem because thereafter the NDMP cloning will not work anymore.

Would setting the "FORCE_REC_AFFINITY" help me on this?