NDMP restore fails.

[brin0019]

I am trying to restore a folder from a emc filer and keep getting this error. The filer backups over the lan to a media server.

This is the job logs.

02/01/2012 09:20:04 - begin reading
02/01/2012 09:20:04 - Info bpbrm (pid=10861) sending media manager msg: CONTINUE RESTORE
02/01/2012 09:20:06 - Error ndmpagent (pid=10881) ndmp_mover_set_record_size failed, status = 9 (NDMP_ILLEGAL_ARGS_ERR)
02/01/2012 09:20:07 - Error ndmpagent (pid=10881) NDMP restore failed from path /root_vdm_4/ccprd_fs01
02/01/2012 09:20:12 - Info ndmpagent (pid=10881) done. status: 114: unimplemented error code 114
02/01/2012 09:20:12 - Info bpbrm (pid=10861) got ERROR 5 from media manager
02/01/2012 09:20:12 - Info bpbrm (pid=10861) killing bpbrm child 10876.
02/01/2012 09:20:13 - restored from image xxxxxxxxxxxxxx_1327558633; restore time: 0:01:08
02/01/2012 09:20:13 - end Restore; elapsed time 0:01:15
Failed to get status code information (2826)
 

This is the ndmp logs from the filer.

2012-02-01 11:20:06: NDMP: 10: Dest IP used by client is:0x8e80160a
2012-02-01 11:20:06: NDMP: 7: (thread ndmp014) Mover init() with Buf==NULL, caller 0x12101b8, callercaller 0x121bc41
2012-02-01 11:20:06: NDMP: 10: Thread ndmp014 got thrdName (ndmp014), sessionId=014.
2012-02-01 11:20:06: NDMP: 10: last message repeated 1 times
2012-02-01 11:20:06: NDMP: 7: < NDMP notify connection version: 4 >
2012-02-01 11:20:06: NDMP: 7: < send NDMP_NOTIFY_CONNECTION_STATUS to client, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) is initialized and ready for NDMP requests.
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < receive CONNECT_OPEN request, protocol_version: 4 >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONNECT_OPEN reply, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: < LOG type: 1, msg_id: 0, entry: The process id for NDMP service is 0xd5170a60, hasAssociatedMsg: 0, associatedMsgSeq: 0 >
2012-02-01 11:20:06: NDMP: 7: < send NDMP_LOG_MESSAGE to client, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < receive CONFIG_GET_SERVER_INFO request >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < vendor_name:EMC2, product_name:CELERRA, revision_number:T.5.6.51.3 >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < support auth_type: NDMP_AUTH_TEXT >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < support auth_type: NDMP_AUTH_MD5 >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONFIG_GET_SERVER_INFO reply, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: < Session 014 (thread ndmp014) receives CONFIG_GET_AUTH_ATTR request, auth_type: NDMP_AUTH_MD5 >
2012-02-01 11:20:06: NDMP: 7: < Session 014 (thread ndmp014) sent CONFIG_GET_AUTH_ATTR reply, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < receive CONNECT_CLIENT_AUTH request, auth_type: NDMP_AUTH_MD5 >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < user: ndmp >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONNECT_CLIENT_AUTH reply, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < receive CONFIG_GET_HOST_INFO request >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < hostname:server_3, os_type:DartOS, os_vers:EMC Celerra File Server.T.5.6.51.3, hostid:abc19
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONFIG_GET_HOST_INFO reply, error: NDMP_NO_ERR >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < received CONFIG_GET_EXT_LIST request >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONFIG_GET_EXT_LIST reply >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < receive CONFIG_GET_CONNECTION_TYPE request >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < support ndmp_addr_type: NDMP_ADDR_LOCAL >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < support ndmp_addr_type: NDMP_ADDR_TCP >
2012-02-01 11:20:06: NDMP: 7: Session 014 (thread ndmp014) < sent CONFIG_GET_CONNECTION_TYPE reply, error: NDMP_NO_ERR >
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) < receive DATA_GET_STATE request >
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) < sent DATA_GET_STATE reply, unsupported=3, error: NDMP_NO_ERR, op: NDMP_DATA_OP_NOACTION, stNDMP_DATA_STATE_IDLE, halt_reason: NDMP_DATA_HALT_NA, bytes_processed=0, est_bytes_remain=0, est_time_remain=0, data_connection_addr=0x0, read_offset=0,_len=0 >
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) < receive DATA_GET_STATE request >
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) < sent DATA_GET_STATE reply, unsupported=3, error: NDMP_NO_ERR, op: NDMP_DATA_OP_NOACTION, stNDMP_DATA_STATE_IDLE, halt_reason: NDMP_DATA_HALT_NA, bytes_processed=0, est_bytes_remain=0, est_time_remain=0, data_connection_addr=0x0, read_offset=0,_len=0 >
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) < receive CONNECT_CLOSE request >
2012-02-01 11:20:07: NDMP: 7: < Thread ndmp014: EOF encountered for this connection in processMessages.>
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) will end the connection (0xd5902604) due to EOF, data and mover will be aborted.
2012-02-01 11:20:07: NDMP: 7: Session 014 (thread ndmp014) ended the session (0xd39cdcac) and connection (0xd5902604).

[Marianne]

Please share more details about your restore steps - selection, restore destination, etc.

NDMP is saying something about 'illegal arguments'?

Please get us the NDMP logs that corresponds with the restore time. Or is there a 2 hour time difference between the filer and the backup server?

Have you tried to select the files in the folder instead of the folder? I have in the the past seen that selection of a folder failed to restore, but that selecting all the files in the folder restored successfully.

Seems that DAR supports individual FILE restore, not directories.

[brin0019]

I have found the solution to the issue.

The restore was trying to use tape drive connected to another filer. I am not too sure why it would do this but disabling the drive and running again allows the restore to use the drives on the media server and not the drive on the filer.

Since the backups are coming over the lan to a media server for backup they are using 256 for buffer size that is set on the media server. Restoring from the same server is not an issue as the buffer size is 256... Since this restore was trying to use drives off another filer for the restore the buffer size on the filer by default is 128 causing the failures.

Problem now is I need to figure out why it would try to use a different filer for the restore and not the media server.

[ayhansevimli]

Change the administration console using for restore !!!

When I use (on my pc that running windows 7 ) java console for restoring NDMP (CIFS) source I recieved same error code 2826.
I use solaris master server in my enviroment.I opened netbackup administration console by using Xmanager program on solaris master server , than I applied same restore steps. There was no problem.

Best regards.