NDMP restores fail with (NDMP_ILLEGAL_ARGS_ERR)

[moez]

Hi,

I'm trying to restore some data from a filer but keep getting the below error. I have tried restoring various different folders to various locations and all fail with the same error. I have disabled/re-enabled the SCSI ports on the filer - ensured detection of tape drives by the filer, killed NDMPD and restarted it all to no avail. Any ideas on what is causing this error?

13:20:56 (274989.001) Restoring from image created 17/01/2009 5:47:24 AM

13:21:08 (274989.001) INF - If Media id XXX123 is not in a robotic library administrative interaction may be required to satisfy this mount request.

13:21:13 (274989.001) INF - Waiting for mount of media id XXX123 on server XXXX01 for reading. 13:21:58 (274989.001) INF - Waiting for positioning of media id XXX123 on server XXXX01 for reading. 13:22:29 (274989.001) INF - Beginning restore from server XXXX01 to client XXXX01.

13:22:35 (274989.001) INF - Restoring NDMP files from /vol/folder1/folder2/ to /vol/folder1/folder2/folder3

13:22:41 (274989.001) DAR enabled

13:22:44 (274989.001) ndmp_data_start_recover failed, status = 9 (NDMP_ILLEGAL_ARGS_ERR)

13:22:48 (274989.001) NDMP restore failed from path /vol/folder1/folder2

13:22:54 (274989.001) Status of restore from image created 17/01/2009 5:47:24 AM = file read failed

13:22:54 (274989.xxx) INF - Status = the restore failed to recover the requested files.

[Michael_Rae]

I have had a similar issue to this myself i believe. If i remember correctly it was due to the restore trying to use a drive that did not have a direct path to the NAS box. If the media server doing the restore has the NDMP drives in use it will try and use another drive in the same library, this will cause this problem.

[Omar_Villa]

run tpconfig -dnh and see if your media server is configured to connect to the filer you are trying to restore too, also check that you have a NDMP drive configured to this filer too, because you can have the access but without a drive configured.

hope this helps.

regards

[Android]

You may also want to confirm that the NDMP device s using a version of the NDMP protocol that is supported by your NetBackup version.  Also that your authentication paramters are set correctly. See the below excerpts from NDMP.org:

NDMP_ILLEGAL_ARGS_ERR  Protocol version suggested by the client is not
                         supported by the server. The client should retry
                         the request with a lower protocol version number.

NDMP_ILLEGAL_ARGS_ERR  Specified authentication method not supported.

[moez]

Thanks for everyone's reply. I've done a bit more investigation and turns out I am able to restore data if I choose individual files and kick off a restore job for each file individualy. It only fails with the NDMP_ILLEGAL_ARGS_ERR when I choose an entire folder (and all it's sub folders) to restore. Having said that I assume I can rule out an authentication or protocol issue.

Unfortunately my work-around won't cut it as you'd imagine it wouldn't be fun kicking off 2000 seperate restore jobs to recover someone's files. So I'm still searching for a solution to this.

thanks for all the help and advice thus far.

[Darren_Dunham]

What kind of machine is your NDMP host?  Have you examined the NDMP logs on that side?  On a netapp filer, you'll need to increase the verbosity to get good ndmp logging.

-- 

Darren

[Methos8]

I'm having this same issue.  I'm restoring to a NetApp filer.  Although NetBackup shows the tape being loaded and read, the ndmp log from the filer seems to suggest it's not finding the tape:

Apr 23 15:05:36 PDT [ndmpd:95]: Error code: NDMP_NO_TAPE_LOADED_ERR
Apr 23 15:05:36 PDT [ndmpd:95]: Device name: nrst0a
Apr 23 15:05:36 PDT [ndmpd:95]: Mode: 0
Apr 23 15:05:36 PDT [ndmpd:95]: IOException: Device cannot be opened. Device may have no tape.
Apr 23 15:05:37 PDT [ndmpd:95]: NDMP message type: NDMP_CONNECT_CLOSE
Apr 23 15:05:37 PDT [ndmpd:95]: NDMP message replysequence: 7
Apr 23 15:05:37 PDT [ndmpd:95]: Message Header:
Apr 23 15:05:37 PDT [ndmpd:95]: Sequence 0
Apr 23 15:05:37 PDT [ndmpd:95]: Timestamp 0
Apr 23 15:05:37 PDT [ndmpd:95]: Msgtype 1
Apr 23 15:05:37 PDT [ndmpd:95]: Method 2306
Apr 23 15:05:37 PDT [ndmpd:95]: ReplySequence 7
Apr 23 15:05:37 PDT [ndmpd:95]: Error NDMP_NO_ERR

[schmaustech]

Does your sysconfig on the Netapp show this particular device as being there?  I have seen it where sometimes the device disappears.

Regards,

Benjamin Schmaus

[Darren_Dunham]

Make sure you line up the logs with the actual restore.  It seems to me that the logs get filled with a lot of "no tape loaded" because it's constantly polling to see if any tapes are in the drive.  Lots of those errors seem to be normal as long as they're not right at the time of the backup.

--
Darren

[Srikanth_Gubbal]

go to Host properties of master server and disable DAR, restart the services or use admincmd/bprdreq -rereadconfig. then tyr restore of folder.

let me know the result,

[Pravs]

you can try disable Directory DAR by using the setting in ndmp.cfg - NDMP_DAR_DIRECTORY_DISABLED Actually, I will say it's NetApp problem. they should be sending us an environmental Variable of "Enhanced_DAR_ENABLED=F" to turn off Directory DAR, but they don't. you need to get filer logs and ndmp logs to see what env variable was passed and what netbackup parsed.