what 5.6 release are you running ?

Also please post the exact error messages from both the Celerra log and the NetBackup log

Pardon the long message. The version is 5.6.38-2

FROM NETBACKUP LOG:
connected; connect time: 0:00:00
Error bptm (pid=1165) NDMP_LOG_ERROR 0 Medium error
Warning bptm (pid=1165) WRN - Cannot backup filepath /mt_javad, NDMP_HALT_INTERNAL_ERROR occurred (path may not exist), skipping
Error bptm (pid=1165) none of the NDMP backups for client cha-nas-nfs1 completed successfully
Error bptm (pid=1165) io_ioctl_ndmp (MTBSF) failed on media id CHA274, drive index 3, return code 7 (NDMP_IO_ERR)(bptm.c.7947)
end writing; write time: 0:32:24


FROM CELERRA LOG:
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) < Backup type: tar >
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: TYPE Value: tar
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: FILESYSTEM Value: /mt_javad
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: PREFIX Value: /mt_javad
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: LEVEL Value: 0
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: HIST Value: y
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) Name: UPDATE Value: y
2008-09-10 14:02:42: NDMP: 4: Session 470 (thread ndmp470) tape drive: c160t0l1, CALLER_ADDRESS 0xfabaac.
2008-09-10 14:02:42: NDMP: 4: Thread ndmp470 uses a set of Pax threads with handle 0.
2008-09-10 14:02:42: NDMP: 6: Thread ndmp470 filterDialect could not be set - prereqs not met [pax_cmd_thread::setFilterDialect]
2008-09-10 14:02:42: NDMP: 4: Thread ndmp470 fsSize of /mt_javad is 90821042176 (../pax_ndmp.cxx: 1173)
2008-09-10 14:35:06: CAM: 3: I/O Error: c160t0l1: CamStatus 0x84 ScsiStatus 0x02 Sense 0x0b/0x00/0x00
2008-09-10 14:35:06: CAM: 3: camFlags 0x91 Addr 0x75f43404 Len 0xfc00 Resid 0x0
2008-09-10 14:35:06: CAM: 3: cdb: 0a 00 00 fc 00 00 00 00 00 00 00 00
2008-09-10 14:35:06: NDMP: 3: Thread nasw00 DartTapeInterface: write error
2008-09-10 14:35:06: NDMP: 3: Session 470 (thread nasw00) Tape write error: No error reported on drive c160t0l1.
2008-09-10 14:35:06: NDMP: 3: Session 470 (thread nasw00 fails in tapeWrite or tapeWriteMsg for drive c160t0l1, numMsg=9, maxMsg=64, n=-1, bcnt=64512, length=64512, fileName=../NdmpdMover.cxx, lineNo=690.
2008-09-10 14:35:06: NDMP: 3: Session 470 (thread nasw00) NdmpdSession::ndmpdApiWrite fails in local or remote wirte msg, moverAddressType=0, mp=0x79051374
2008-09-10 14:35:06: NDMP: 3: Thread bkup084 Medium error
2008-09-10 14:35:06: NDMP: 3: < LOG type: 2, msg_id: 0, entry: Medium error, hasAssociatedMsg: 0, associatedMsgSeq: 0 >
2008-09-10 14:35:06: NDMP: 4: Thread bkup084 Write failed on archive volume 1
2008-09-10 14:35:06: NDMP: 4: < LOG type: 3, msg_id: 0, entry: Write failed on archive volume 1, hasAssociatedMsg: 0, associatedMsgSeq: 0 >
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 Backup root directory: /mt_javad
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 write count 69082094592
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 time used 1947 sec
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 write rate 34644 KB/Sec
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 read rate 0 KB/Sec
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 average file size: 152KB
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 dir or 0 size file processed: 79855
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 1B -- 8KB size file processed: 245197
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 8KB -- 16KB size file processed: 41373
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 16KB -- 32KB size file processed: 24528
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 32KB -- 64KB size file processed: 16708
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 64KB -- 1MB size file processed: 29429
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 1MB -- 32MB size file processed: 4501
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 32MB -- 1GB size file processed: 317
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 1G more size file processed: 4
2008-09-10 14:35:06: NDMP: 6: Thread nasa00 server_archive: emctar vol 1, 441912 files, 0 bytes read, 69082094592 bytes written
2008-09-10 14:35:06: NDMP: 4: Thread nasa00 cifs_ea new count: 21513455, delete count: 21513455, freeListCount: 128
2008-09-10 14:35:06: PAX: 4: Thread nasa00 used these threads for this job: bkup033 bkup081 bkup008 bkup001 bkup023 bkup060 bkup105 bkup034 bkup030 bkup040 bkup042 bkup039 bkup028 bkup084 bkup015 bkup009 bkup106 bkup013 bkup029 bkup011 bkup112 bkup032 bkup091 bkup126 bkup090 bkup055 bkup043 bkup005 bkup087 bkup104 bkup076 bkup080 bkup098 bkup092 bkup007 bkup088 bkup017 bkup035 bkup117 bkup020 bkup114 bkup083 bkup066 bkup016 bkup120 bkup070 bkup024 bkup062 bkup078 bkup018 bkup073 bkup054 bkup067 bkup050 bkup101 bkup100 bkup012 bkup046 bkup107 bkup097 bkup010 bkup061 bkup041 bkup086
2008-09-10 14:35:06: CAM: 3: I/O Error: c160t0l1: CamStatus 0x84 ScsiStatus 0x02 Sense 0x03/0x52/0x00
2008-09-10 14:35:06: CAM: 3: camFlags 0xc1 Addr 0xffffffff Len 0x0 Resid 0x0
2008-09-10 14:35:06: CAM: 3: cdb: 10 00 00 00 01 00 00 00 00 00 00 00

5.6.38-2

connected; connect time: 0:00:00
Error bptm (pid=1165) NDMP_LOG_ERROR 0 Medium error
Warning bptm (pid=1165) WRN - Cannot backup filepath /mt_javad, NDMP_HALT_INTERNAL_ERROR occurred (path may not exist), skipping
Error bptm (pid=1165) none of the NDMP backups for client cha-nas-nfs1 completed successfully
Error bptm (pid=1165) io_ioctl_ndmp (MTBSF) failed on media id CHA274, drive index 3, return code 7 (NDMP_IO_ERR)(bptm.c.7947)
end writing; write time: 0:32:24

Hi,
Just wondering are you using the automatic checkpoints by setting the "set snapsure =yes" setting on ur netbackup policy? Because we are also running 5.6 and doing NDMP backups from our DR NS40 however we discovered a bug in 5.6 where when it mounts the automatically created checkpoint the PFS becomes unmounted. What we had to do to fix it was take out the set snapsure =yes and just create a manual checkpoint schedule to overwrite itself each night just before the backup runs then put this check point name in the backup selection..

might be worth a try in your case..

Packetboy.

Ah, I see - you're doing NDMP backup with integrated checkpoints on the remote side of an IP RepV2

This is indeed fixed and waiting for a maintenance release

You got a knowledgebase entry for that now as well - emc197062

Rainer
P.S.: no - I dont know which release or when - you have to ask that through your escalation contacts

@packetboy - do you have a service request number for that ?

Hi Rainer,
Sure do its 24541004 its still open and says its with engineering still awaiting release..

thanks,

Packetboy

Hey thanks heaps Rainer.. one thing can you link direct to the KB article? I cant find it when I search in powerlink?

Thanks,

Packetboy

This look similar to a fix we did in 5.6.38

I suggest to open a case with Symantec and EMC support - maybe NetBackup needs a patch to handle the error correctly

Symptom Description

After upgrading to Celerra Network Server version5.5, NDMP backups were automatically aborting full /
incremental backups on backing up 512kb with medium errors. The following error is seen in the server_log:

2007-08-23 14:57:14: CAM: 3: cdb: 0a 00 00 f0 00 00 00 00 00 00
00 00 | SCSI ERR ON #4800 camstat 0x84 bstat DeviceError |
scsi stat.skey.ascq 0x02.03.3b00 Check condition | skey [ ]
Medium error _____ ascq Sequential positioning error
2007-08-23 14:57:14: NDMP: 3: DartTapeInterface: write error
2007-08-23 14:57:14: NDMP: 3: Tape write error: No error
reported .

In prior releases, Celerra sent an error of weof failed to the NDMP host when this event
occurred, but no longer does. NetWorker relies on this error, and when it didn't receive the error, it could
not mark the volume to full and skip it. Instead, backups failed, and Networker attempted to reuse the
bad tapes in future backups.

Fix
Media errors are now reported with ndmp_mtio commands and Legato Networker marks tapes as "full" as expected.

Service Request No. 19414442
Fixed in version 5.6.38.0

I guess its because its still in draft status. here's what it says:

ID: emc197062

Fact Celerra

Fact Feature: Celerra Replication V2

Fact NDMP Backups

Fact Feature: SnapSure Checkpoint

Symptom Backup of the destination side of Replication V2 with auto checkpoint / snapsure enabled fails

Symptom Server_log reports "SnapSure file system creation fails"

Symptom From the server_log and sys_log you can see that the checkpoint gets created.

Symptom The checkpoint goes inactive immediately as the FS gets unmounted - through freeze/thaw action of replication.

Symptom fs_ckpt FS -l output while the backup is running you can see that the checkpoint is created

Symptom If you do a server_mount server_x the FS looks mounted but if you do a server_df server_x the FS is not in the output

Fix The fix is still under development - Please contact your local EMC support center quoting this Knowledgebase Solution to get an update on the development of this fix

The workaround is to create a manual checkpoint of the remote fs and backup that checkpoint.