Greetings,
I am trying to restore an individual file from a TSM NDMP backup.
This is a 3.7TB filesystem, so my predecessor didn't think it was worth the
disk space to throw at a TOC backup, so I have to restore using 'restore
node'.
TSM version 5.3.4, running under AIX 5.3ML1. Here is the console from my
first restore attempt:
tsm: MDCTSM00>restore node cx02_dm2 /fs_dst
filelist=/dst/csdforms/CSD/UpdateCSD
s.vbs pitd=10/10/2006 pittime=23:00:00
ANR1059I Selective restore of NAS node CX02_DM2, file system /fs_dst,
started
as process 933 by administrator SCHNJD. Specified files and/or directory
trees
will be restored to destination /fs_dst.
ANS8003I Process number 933 started.
tsm: MDCTSM00>q proc
Process Process Description Status
Number
-------- --------------------
-------------------------------------------------
933 Restore NAS NAS Node CX02_DM2, Source /fs_dst, Destination
(selective) /fs_dst, Elapsed Time 22 seconds. Waiting for
mount of input volume 010259 (21 seconds).
tsm: MDCTSM00>q req
ANR8352I Requests outstanding:
ANR8308I 003: NAS volume 010259 is required for use in library IBM3584;
CHECKIN
LIBVOLUME required within 179 minutes.
tsm: MDCTSM00>q nasbackup cx02_dm2 *
Node Name Filespace Object Type Object Size Creation Has Table of
Name (MB) Date Contents
(TOC)?
--------------- ----------- ----------- ----------- ---------- -------------
CX02_DM2 /fs_dst Differenti- 61,393.7 10/05/2006 No
al Image 23:00:01
CX02_DM2 /fs_dst Differenti- 67,050.3 10/06/2006 No
al Image 23:00:00
CX02_DM2 /fs_dst Full Image 3,711,751.2 10/07/2006 No
18:00:00
CX02_DM2 /fs_dst Differenti- 3,736.2 10/08/2006 No
al Image 23:00:01
CX02_DM2 /fs_dst Differenti- 8,003.7 10/09/2006 No
al Image 23:00:01
CX02_DM2 /fs_dst Differenti- 15,614.2 10/10/2006 No
al Image 23:00:00
CX02_DM2 /fs_dst Differenti- 21,198.5 10/11/2006 No
al Image 23:00:00
I mounted 010259 as it asked for it, and after it mounted the tape it ran
for over an hour and then produced:
10/12/06 14:28:03 ANR9999D ndlog.c(105): ThreadId<40> NDMPLOG: E Read
failed
on archive volume 1 (SESSION: 128391, PROCESS: 933)
10/12/06 14:28:03 ANR9999D ThreadId<40> issued message 9999 from:
<-0x000000010001c168 outDiagf <-0x00000001002e6c3c
spiOutMsg <-0x09000000005896f0 ndmpLogMessage
<-0x090000-
0000580608 ndmp_process_messages <-0x09000000005810e4
ndmpProcessRequests <-0x09000000005886a4 ndmpPoll
<-0x09000000005771d4 spiRtrvSeg <-0x000000010040e728
AsReadSegRemote <-0x00000001004ae1e0 ssRtrvRemote
<-0x0000000100497b10 AfRtrvRemoteThread
<-0x000000010000-
e9dc StartThread <-0x09000000003402dc _pthread_body
(SESSION: 128391, PROCESS: 933)
10/12/06 14:28:08 ANR1104E NAS Restore process 933 terminated - NDMP
session
errors encountered. (SESSION: 128391, PROCESS: 933)
10/12/06 14:28:08 ANR0985I Process 933 for RESTORE NAS (SELECTIVE)
running
in the BACKGROUND completed with completion state
FAILURE
at 14:28:08. (SESSION: 128391, PROCESS: 933)
10/12/06 14:28:08 ANR0514I Session 128391 closed volume 010259.
(SESSION:
128391)
10/12/06 14:28:37 ANR8468I NAS volume 010259 dismounted from drive
DRIVE262
(c112t0l0) in library IBM3584. (SESSION: 128391,
PROCESS:
933)
I looked back through the activity log, and it looks like 010259 is the
first tape in the previous full backup for this filesystem.
On the Celerra, we get the following:
2006-10-12 14:17:47: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator check
condition: loopID = 3., SK = 06, ASC/Q = 2800
2006-10-12 14:17:47: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator check
condition: loopID = 3., SK = 02, ASC/Q = 3a00
2006-10-12 14:17:51: DRIVERS: 4: last message repeated 1 times
2006-10-12 14:17:51: DRIVERS: 4: FCDMTL 2 [1.4.1] Suppressed all but 2 out
of 4 identical log entries
2006-10-12 14:17:51: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator check
condition: loopID = 3., SK = 02, ASC/Q = 0401
2006-10-12 14:18:03: DRIVERS: 4: last message repeated 1 times
2006-10-12 14:18:03: CAM: 3: check_timeout (112) : c112t0l0 req bfba5090
timeout (10 sec 39236029860 cyc)
func 01 flags 41 stat 00 cdb 380a4c1a000000000000
2006-10-12 14:18:03: FCP: 3: c112t0l0: abort called
2006-10-12 14:18:03: DRIVERS: 4: FCDMTL 2 [1.4.1] Suppressed all but 2 out
of 4 identical log entries
2006-10-12 14:18:03: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator command error:
loop_id = 3., SCSI status = 0, EXM status = 12
2006-10-12 14:18:03: CAM: 3: 6: Attention: Scsi-112 hba is AGAIN NORMAL.
|_SCSI ERR ON #11200 camstat 0x04 bstat DeviceError
2006-10-12 14:18:07: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator check
condition: loopID = 3., SK = 06, ASC/Q = 2800
2006-10-12 14:18:15: DRIVERS: 4: FCDMTL 2 [1.4.1] Initiator check
condition: loopID = 3., SK = 00, ASC/Q = 0001
2006-10-12 14:18:41: NDMP: 4: Name: FILESYSTEM Value: /fs_dst
2006-10-12 14:18:41: NDMP: 4: Name: LEVEL Value: 0
2006-10-12 14:18:41: NDMP: 4: Name: UPDATE Value: Y
2006-10-12 14:18:41: NDMP: 4: Name: NON_QUOTA_TREE Value: N
2006-10-12 14:18:41: NDMP: 4: Name: NO_ACLS Value: N
2006-10-12 14:18:41: NDMP: 4: Name: HIST Value: N
2006-10-12 14:18:41: NDMP: 4: Name: DIRECT Value: N
2006-10-12 14:18:41: NDMP: 4: Name: EXTRACT Value: Y
2006-10-12 14:18:41: NDMP: 4: Name: EXTRACT_ACL Value: Y
2006-10-12 14:18:41: NDMP: 4: Name: DEBUG Value: N
2006-10-12 14:18:41: NDMP: 4: Name: VERBOSE Value: N
2006-10-12 14:18:41: NDMP: 4: Name: REPLICATE Value: N
2006-10-12 14:18:41: NDMP: 4: Name: FORCE Value: N
2006-10-12 14:18:41: NDMP: 4: Name: NOWRITE Value: N
2006-10-12 14:18:41: NDMP: 4: Name: LIST Value: N
2006-10-12 14:18:41: NDMP: 4: Name: LIST_TREES Value: N
2006-10-12 14:18:41: NDMP: 4: Name: DMP_NAME Value: TSM /fs_dst
2006-10-12 14:18:41: NDMP: 4: Name: DATA_BLOCK_SIZE Value: 0
2006-10-12 14:18:41: NDMP: 4: Name: TYPE Value: DUMP
2006-10-12 14:18:41: NDMP: 4: < Backup type: dump >
Note that the tape drive in this case is c112t0l0, known to TSM as DRIVE262.
My second attempt was to go back and specify the latest full as the
point-in-time:
tsm: MDCTSM00>q nasbackup cx02_dm2 *
Node Name Filespace Object Type Object Size Creation Has Table of
Name (MB) Date Contents
(TOC)?
--------------- ----------- ----------- ----------- ---------- -------------
CX02_DM2 /fs_dst Differenti- 67,050.3 10/06/2006 No
al Image 23:00:00
CX02_DM2 /fs_dst Full Image 3,711,751.2 10/07/2006 No
18:00:00
CX02_DM2 /fs_dst Differenti- 3,736.2 10/08/2006 No
al Image 23:00:01
CX02_DM2 /fs_dst Differenti- 8,003.7 10/09/2006 No
al Image 23:00:01
CX02_DM2 /fs_dst Differenti- 15,614.2 10/10/2006 No
al Image 23:00:00
CX02_DM2 /fs_dst Differenti- 21,198.5 10/11/2006 No
al Image 23:00:00
tsm: MDCTSM00>restore node cx02_dm2 /fs_dst
filelist=/dst/csdforms/CSD/UpdateCSD
s.vbs pitd=10/07/2006 pittime=18:00:00
ANR1059I Selective restore of NAS node CX02_DM2, file system /fs_dst,
started
as process 971 by administrator SCHNJD. Specified files and/or directory
trees
will be restored to destination /fs_dst.
ANS8003I Process number 971 started.
tsm: MDCTSM00>
I am trying to restore a file which is on the full backup for 10/07/2006
18:00:00. Am I doing the right thing, then, when I specify that
point-in-time for the restore? Or do I have to specify a point-in-time in
the future? The reason I ask is because I it asked to mount tape 010404,
but that tape was not used in the full backup on that date. Here are the
tapes that it used for that full:
010259
010261
030260
030774
030901
030925
030961
031034
031054
031056
031088
031142
Either way, we got the same error as before:
10/13/06 10:49:13 ANR9999D ndlog.c(105): ThreadId<46> NDMPLOG: E Read
failed
on archive volume 1 (SESSION: 132921, PROCESS: 971)
10/13/06 10:49:13 ANR9999D ThreadId<46> issued message 9999 from:
<-0x000000010001c168 outDiagf <-0x00000001002e6c3c
spiOutMsg <-0x09000000005896f0 ndmpLogMessage
<-0x090000-
0000580608 ndmp_process_messages <-0x09000000005810e4
ndmpProcessRequests <-0x09000000005886a4 ndmpPoll
<-0x09000000005771d4 spiRtrvSeg <-0x000000010040e728
AsReadSegRemote <-0x00000001004ae1e0 ssRtrvRemote
<-0x0000000100497b10 AfRtrvRemoteThread
<-0x000000010000-
e9dc StartThread <-0x09000000003402dc _pthread_body
(SESSION: 132921, PROCESS: 971)
10/13/06 10:49:18 ANR1104E NAS Restore process 971 terminated - NDMP
session
errors encountered. (SESSION: 132921, PROCESS: 971)
10/13/06 10:49:18 ANR0985I Process 971 for RESTORE NAS (SELECTIVE)
running
in the BACKGROUND completed with completion state
FAILURE
at 10:49:18. (SESSION: 132921, PROCESS: 971)
10/13/06 10:49:18 ANR0514I Session 132921 closed volume 010404.
(SESSION:
132921)
10/13/06 10:49:46 ANR8468I NAS volume 010404 dismounted from drive
DRIVE263
(c160t0l0) in library IBM3584. (SESSION: 132921,
PROCESS:
971)
Note that the tape drive in this case is c160t0l0, known to TSM as DRIVE263.
So I am getting the same problem on two different tape backups on two
different days, and two different tapes, restoring from two different tape
drives.
I would be eternally grateful for any suggestions. I have opened up a PMR
for this, but they have not yet responded.
Best Regards,
John D. Schneider
Senior Systems Administrator - Storage
Sisters of Mercy Health System
3637 S. Geyer Rd.
St. Louis, MO 63127
Email: schnjd AT stlo.mercy DOT net
Phone: 314-364-3150 Cell: 314-486-2359
|