I already had a horible problem w/ ndmp copy included w/ ontapp : there is an option to preserve /etc but in fact, it didn't did that, it tried to replace some files in /etc and i was able to find 0 byte sized files in /etc
it was on an old 6.X release i thought netapp was able to correct this problem, i was wrong
make sure the ndmpcopy didn't write in a currently used /etc directory because the dest filer needs its /etc directory being untouched
::trick::* if you want to get all the log, lauch the ndmpcopy throught rsh and pipe it throught a * tee * or redirect it to a file : rsh myfiler ndmpcopy... > mytrace.log rsh myfiler ndmpcopy... | tee mytrace.log
Darragh, Stephen J wrote:
Hello,
I am running 6.4.1 on both the F740 (source) and the F825 (target) and ran into a strange ndmpcopy problem
From my unix admin host, I saved a copy of the etc directory before I started the copy. sungod# tar cf - . | (cd ../etc.save; tar xfBp -) sungod# pwd /mnt2/etc sungod# df -k . Filesystem kbytes used avail capacity Mounted on 129.86.9.201:/vol/nhq01 601529792 60321248 541208544 11% /mnt2
sansrvnas01> ndmpcopy -da root:csc01admin -l 0 /vol/nhq01 sansrvnas10:/vol/nhq01 Ndmpcopy: Starting copy [ 10 ] ... Ndmpcopy: sansrvnas01: Notify: Connection established Ndmpcopy: sansrvnas10: Notify: Connection established Ndmpcopy: sansrvnas01: Connect: Authentication successful Ndmpcopy: sansrvnas10: Connect: Authentication successful Ndmpcopy: sansrvnas01: Log: DUMP: creating "/vol/nhq01/../snapshot_for_backup.149" snapshot. Ndmpcopy: sansrvnas01: Log: DUMP: Using Full Volume Dump Ndmpcopy: sansrvnas01: Log: DUMP: Date of this level 0 dump: Thu Jul 3 12:47:51 2003. Ndmpcopy: sansrvnas01: Log: DUMP: Date of last level 0 dump: the epoch. Ndmpcopy: sansrvnas01: Log: DUMP: Dumping /vol/nhq01 to NDMP connection Ndmpcopy: sansrvnas01: Log: DUMP: mapping (Pass I)[regular files] Ndmpcopy: sansrvnas01: Log: DUMP: mapping (Pass II)[directories] Ndmpcopy: sansrvnas01: Log: DUMP: estimated 301180237 KB. Ndmpcopy: sansrvnas01: Log: DUMP: dumping (Pass III) [directories] Ndmpcopy: sansrvnas10: Log: RESTORE: Thu Jul 3 16:55:49 2003: Begin level 0 restore Ndmpcopy: sansrvnas10: Log: RESTORE: Thu Jul 3 16:55:49 2003: Reading directories from the backup Ndmpcopy: sansrvnas10: Log: RESTORE: Thu Jul 3 16:59:50 2003 : We have read 121799 KB from the backup. Ndmpcopy: sansrvnas01: Log: DUMP: Thu Jul 3 12:58:02 2003 : We have written 122225 KB.
****** This job finished and I could still get to the /etc directory from my unix admin host.
sansrvnas01> ndmpcopy -da root:csc01admin -l 1 /vol/nhq01 sansrvnas10:/vol/nhq01 Ndmpcopy: Starting copy [ 11 ] ... Ndmpcopy: sansrvnas01: Notify: Connection established Ndmpcopy: sansrvnas10: Notify: Connection established Ndmpcopy: sansrvnas01: Connect: Authentication successful Ndmpcopy: sansrvnas10: Connect: Authentication successful Ndmpcopy: sansrvnas01: Log: DUMP: creating "/vol/nhq01/../snapshot_for_backup.158" snapshot. Ndmpcopy: sansrvnas01: Log: DUMP: Using subtree dump Ndmpcopy: sansrvnas01: Log: DUMP: Date of this level 1 dump: Sun Jul 6 15:45:46 2003. Ndmpcopy: sansrvnas01: Log: DUMP: Date of last level 0 dump: Thu Jul 3 12:47:51 2003. Ndmpcopy: sansrvnas01: Log: DUMP: Dumping /vol/nhq01 to NDMP connection Ndmpcopy: sansrvnas01: Log: DUMP: mapping (Pass I)[regular files]
The login times out at some point, so I can't see any error. I cannot get to the etc directory, its empty. I can't see my mount point any longer and the etc.save directory is also wiped out. I have an open case with NetApp [NETAPP LOG # 530454], but want to see if anyone else has seen this, or if I am doing something wrong from the command line. Also, I was not able to pull from the sansrvnas10 (F825), only push from sansrvnas01 (F740).
I looked at ndmpcopy man pages and it says you must use the -f flag if you wanted the /etc directory copied, which I don't because they are not the same type system, but even that would not have created a empty etc directory.
Suggestions?
Stephen Darragh