toasters May 2001

toasters@lists.teaparty.net

173 participants
257 discussions

RE: Question on "vol copy"
by Smith, Gil 24 May '01

24 May '01

> Do any of you know if the "vol copy" command can be used in > this case. Can I specify that the destination is actually a qtree? I would recommend you use ndmpcopy for this. Here is some info and an example: Here is the "ndmpcopy" info for migrating data between filers. It uses dump/restore and can be used for smaller paths (not entire volumes). You have to make sure that each filer trusts each other by making etc/hosts.equiv entries. There is also an actual test level 0 and level 9 below. It is a good tool to know how to use/get working. You can get a binary for ndmpcopy via anonymous ftp from our ftp.netapp.com as below or www.ndmp.org and compile it. ftp> cd ndmpcopy 250 "/pub/netapp/upload/ndmpcopy" is new cwd. ftp> dir 200 PORT command successful. 150 Opening ASCII mode data connection for /bin/ls. drwxrwxr-x 2 ftpuser ftpusers 4096 Apr 17 2000 aix drwxrwxr-x 2 ftpuser ftpusers 4096 Mar 29 2000 linux drwxrwxr-x 2 ftpuser ftpusers 4096 Oct 5 08:54 solaris drwxrwxr-x 2 ftpuser ftpusers 4096 Oct 5 06:52 solaris2.7 Using ndmpcopy for data migration Use the latest version of ndmpcopy. Version 1.2 and later has incremental support that works, and you no longer need to specify a password on the command line, it is promptable now. Refer to ndmpcopy documentation for client installation and command usage. Example using ndmpcopy Version 1.2: Set up the destination volume where the date will be copied to. Use ndmpcopy to do a level 0 dump: Example: ndmpcopy source:path_to_volume destination:path_to_volume -level 0 -dpass After the level 0 is done, a level 1 ndmpcopy may done. This will copy only the data that changed since the level 0: ndmpcopy source:path_to_volume destination:path_to_volume -level 1 -dpass When ready to use the new volume: Turn off NFS & CIFS on the source by executing nfs off and cifs terminate. Do a final incremental to copy changes since the last ndmpcopy: ndmpcopy source:path_to_volume destination:path_to_volume -level 9 -dpass Configure the volume for production use. This process maintains performance by keeping the data striped across all disks in the volume, and you would probably rather wait the extra few minutes of downtime for the ndmpcopy to complete and maintain performance, than lose a lot of performance to get it done quickly. Be sure to use the latest recommended Data ONTAP OS. [example from lab below] unix# ndmpcopy f760cl1-rtp:/vol/vol0/home f820-rtp:/vol/vol0/home -level 0 -dpass Source password: Destination password: Connecting to f760cl1-rtp. Connecting to f820-rtp. f760cl1-rtp: CONNECT: Connection established. f820-rtp: CONNECT: Connection established. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: creating "/vol/vol0/../snapshot_for_ba ckup.1" snapshot. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Using subtree dump f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Date of this level 0 dump: Tue Mar 27 14:32:47 2001. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Date of last level 0 dump: the epoch. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Dumping /vol/vol0/home/ to NDMP connec tion f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: mapping (Pass I)[regular files] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: mapping (Pass II)[directories] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: estimated 146499 KB. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass III) [directories] f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:26:48 2001: Begin level 0 r estore f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:26:48 2001: Reading directo ries from the backup f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass IV) [regular files] f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:26:48 2001: Creating files and directories. f820-rtp: LOG: RESTORE: f820-rtp: LOG: Warning: cannot create directory oracle: File exists f820-rtp: LOG: RESTORE: f820-rtp: LOG: Warning: cannot create directory test: Fi le exists f820-rtp: LOG: RESTORE: f820-rtp: LOG: Warning: cannot create directory dmj: Fil e exists f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:26:48 2001: Writing data to files. f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:27:08 2001: Restoring NT AC Ls. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass V) [ACLs] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: 146901 KB f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: DUMP IS DONE f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Deleting "/vol/vol0/../snapshot_for_ba ckup.1" snapshot. f820-rtp: LOG: RESTORE: f820-rtp: LOG: RESTORE IS DONE f820-rtp: HALT: The operation was successful! Waiting for f760cl1-rtp to halt too. The transfer was successful, but the source filer hasn't halted yet. Elapsed time: 0 hours, 0 minutes, 49 seconds. f760cl1-rtp: HALT: The operation was successful! The transfer is complete. Elapsed time: 0 hours, 0 minutes, 50 seconds. ------------------ now the incremental ------------------------- root@ultra5-rtp%[/usr/local/bin]>./ndmpcopy f760cl1-rtp:/vol/vol0/home f820-rtp: /vol/vol0/home -level 9 -dpass Source password: Destination password: Connecting to f760cl1-rtp. Connecting to f820-rtp. f760cl1-rtp: CONNECT: Connection established. f820-rtp: CONNECT: Connection established. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: creating "/vol/vol0/../snapshot_for_ba ckup.3" snapshot. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Using subtree dump f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Date of this level 9 dump: Tue Mar 27 14:41:38 2001. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Date of last level 0 dump: Tue Mar 27 14:38:20 2001. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Dumping /vol/vol0/home/ to NDMP connec tion f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: mapping (Pass I)[regular files] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: mapping (Pass II)[directories] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: estimated 10274 KB. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass III) [directories] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass IV) [regular files] f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:35:38 2001: Begin increment al restore f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:35:39 2001: Reading directo ries from the backup f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:35:39 2001: Creating new fi les and directories. f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:35:39 2001: Writing data to files. f820-rtp: LOG: RESTORE: f820-rtp: LOG: Tue Mar 27 19:35:41 2001: Restoring NT AC Ls. f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: dumping (Pass V) [ACLs] f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: 10681 KB f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: DUMP IS DONE f760cl1-rtp: LOG: DUMP: f760cl1-rtp: LOG: Deleting "/vol/vol0/../snapshot_for_ba ckup.3" snapshot. f820-rtp: LOG: RESTORE: f820-rtp: LOG: RESTORE IS DONE f820-rtp: HALT: The operation was successful! Waiting for f760cl1-rtp to halt too. The transfer was successful, but the source filer hasn't halted yet. Elapsed time: 0 hours, 0 minutes, 45 seconds. f760cl1-rtp: HALT: The operation was successful! The transfer is complete. Elapsed time: 0 hours, 0 minutes, 46 seconds.

1 0

Re: Question on "vol copy"
by dmaloy＠atomaintl.com 24 May '01

24 May '01

Allen, The method that I have used in the past which I believe works better than vol copy is the tar command: mount both filesystems to your adminhost: mount tornado:/vol/users /tornado/vol/users mount tempest:/vol/vol0/users /tempest/vol/vol0/users cd to /tornado/vol/users and issue the tar command: tar -cf - . | (cd /tempest/vol/vol0/users ; tar -xvpf -) Doing this should also defrag the filesystems which I don't believe vol copy does...Someone can correct me here if I fell asleep during that part of the class. Don Maloy Intier Automotive Sr. CAD Administrator 19888 Haggerty Road Livonia, MI 48152 734.432.4418 "Allen, Pat" <pat(a)mbari.or To: "Toasters (E-mail)" <toasters(a)mathworks.com> g> cc: (bcc: Don Maloy/PDD/DORTEC) Subject: Question on "vol copy" 05/24/01 02:22 PM Hi all, I have an existing 740 (tornado) with two volumes - vol0 and users. I have a new 740 (tempest) with just vol0 and a qtree named "users". I want to copy all the user directories from torando:/vol/users to tempest:/vol/vol0/users in the easiest and quickest method possible. Both of the filers are connected via gig-ethernet to our main switch so using a filer-to-filer transfer would be the fastest. Do any of you know if the "vol copy" command can be used in this case. Can I specify that the destination is actually a qtree? The documentation for the vol command also says that the destination volume must be offline but that the root volume can't be taken offline. Is this a catch-22? Thanks for the help! --- Pat Allen (pat(a)mbari.org) Monterey Bay Aquarium Research Institute (MBARI) 7700 Sandholdt Rd, Moss Landing, CA 95039 (voice) 831-775-1724; (fax) 831-775-1620

2 1

RE: Question on "vol copy"
by Mohler, Jeff 24 May '01

24 May '01

'vol copy' can only be used to migrate data between two separate volumes, you cannot dump a remote volume into a Qtree. What you can do, is use ndmpcopy to migrate data from your remote users volume, to the local users Qtree. That would be as quick, or quicker in some cases than vol copy. Your last resort would be using cp with a unix host in the middle..or an NT solution to copy it all. The reason the destination volume cant be a root volume, is that you need the root volume to boot from..so therefore it cant be offline to be used as a vol copy or snapmirror destination volume. -----Original Message----- From: Allen, Pat [mailto:pat@mbari.org] Sent: Thursday, May 24, 2001 11:22 AM To: Toasters (E-mail) Subject: Question on "vol copy" Hi all, I have an existing 740 (tornado) with two volumes - vol0 and users. I have a new 740 (tempest) with just vol0 and a qtree named "users". I want to copy all the user directories from torando:/vol/users to tempest:/vol/vol0/users in the easiest and quickest method possible. Both of the filers are connected via gig-ethernet to our main switch so using a filer-to-filer transfer would be the fastest. Do any of you know if the "vol copy" command can be used in this case. Can I specify that the destination is actually a qtree? The documentation for the vol command also says that the destination volume must be offline but that the root volume can't be taken offline. Is this a catch-22? Thanks for the help! --- Pat Allen (pat(a)mbari.org) Monterey Bay Aquarium Research Institute (MBARI) 7700 Sandholdt Rd, Moss Landing, CA 95039 (voice) 831-775-1724; (fax) 831-775-1620

1 0

Support of SQL7 on filers pulled by MS?
by Lyons, Jim 24 May '01

24 May '01

MS KB article Q196904 "Support for Network Database Files," which used to detail the manner in which MS would support SQL7 on filers, has disappeared from the MS website. Does anyone know if MS has pulled support for this? Thought I would ask here before wading into MS support... Thanks, Jim

2 1

Question on "vol copy"
by Allen, Pat 24 May '01

24 May '01

Hi all, I have an existing 740 (tornado) with two volumes - vol0 and users. I have a new 740 (tempest) with just vol0 and a qtree named "users". I want to copy all the user directories from torando:/vol/users to tempest:/vol/vol0/users in the easiest and quickest method possible. Both of the filers are connected via gig-ethernet to our main switch so using a filer-to-filer transfer would be the fastest. Do any of you know if the "vol copy" command can be used in this case. Can I specify that the destination is actually a qtree? The documentation for the vol command also says that the destination volume must be offline but that the root volume can't be taken offline. Is this a catch-22? Thanks for the help! --- Pat Allen (pat(a)mbari.org) Monterey Bay Aquarium Research Institute (MBARI) 7700 Sandholdt Rd, Moss Landing, CA 95039 (voice) 831-775-1724; (fax) 831-775-1620

1 0

RE: Appending backups to tapes
by Fox, Adam 24 May '01

24 May '01

I guess I mis-read your question. To *not* append a backup, use a rewinding device. That will automatically rewind the tape at the end of the backup and then you're done. -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com > -----Original Message----- > From: Chris Thompson [mailto:cet1@cus.cam.ac.uk] > Sent: Thursday, May 24, 2001 9:41 AM > To: john.f.mclaughlin(a)lmco.com > Cc: toasters(a)mathworks.com > Subject: Re: Appending backups to tapes > > > "John F. McLaughlin" <john.f.mclaughlin(a)lmco.com> writes > > > > I would like to know if it is possible not to append a backup to a > > previous backup tape. > > Possible *not* to append to a previous tape? Surely: take the > tape away > and lock it up in a safe where the filer can't work its evil ways with > it... :) > > Can you clarify what you mean here? Are you having some problem with > dumps being appended to a tape when you intended to use a new one? > Are you using basic dump/restore commands, or an NDMP > product, or what? > Manually loaded tape drive, or a stacker, or an all-singing > all-dancing > robot? etc. ... > > Chris Thompson University of Cambridge > Computing Service, > Email: cet1(a)ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QG, > Phone: +44 1223 334715 United Kingdom. >

1 0

Appending backups to tapes
by John F. McLaughlin 24 May '01

24 May '01

I would like to know if it is possible not to append a backup to a previous backup tape.

2 1

RE: Appending backups to tapes
by Fox, Adam 24 May '01

24 May '01

It should be. It kind of depends on the software you are using. If you're just using our dump and mt commands on the filer, then just use non-rewinding tape devices and dump again. Then when you do finally rewind you can use mt to fast forward past a specific number of EOFs. Other backup software may have this feature built-in and I can't speak for most of them, but at least natively it's possible. -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com > -----Original Message----- > From: John F. McLaughlin [mailto:john.f.mclaughlin@lmco.com] > Sent: Thursday, May 24, 2001 6:13 AM > To: toasters(a)mathworks.com > Subject: Appending backups to tapes > > > I would like to know if it is possible not to append a backup to a > previous > backup tape. >

1 0

RE: interface fun
by Fox, Adam 23 May '01

23 May '01

I need to retract this statement. I was working off of outdated info. This does *not* work. However, the on-board on the F840 does support VLAN tagging (which I think was another one of your questions). I researched this while I was verifying my first response. Sorry about that. Was working off of faulty memory cells. :) -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com > -----Original Message----- > From: Fox, Adam > Sent: Wednesday, May 23, 2001 5:22 PM > To: 'kevin graham'; toasters(a)mathworks.com > Subject: RE: interface fun > > > > > > Since I don't have the extra interfaces to play with, does > > anyone know if > > a VIF made up of vlan'able interfaces can also be vlaned? > > Rolling all of > > our interfaces into a VIF then VLAN'ing that would be > beautiful, but I > > don't want to go to the trouble having someone persue it if its not > > currently feasible. > > This is in fact true. You can create a big vif and then use vlan > tagging to get up to 4 vlans per vif. This requires at least ONTAP > 6.1. > > -- Adam Fox > NetApp Professional Services, NC > adamfox(a)netapp.com > >

1 0

NetApp F720 -- Password recovery
by Rafi Khardalian 23 May '01

23 May '01

Anyone know how to reset the password and configuration on an F720, without physically having the boot disks? In the past, I called NetApp support with the same problem, at which point I was given instructions on how to emulate a floppy boot. Unfortunately I misplaced this valuable piece of information. The only part I recall is setting a variable at the "ok"/bios prompt named something like ?floppy-boot?. Help would be greatly appreciated. Thanks. Rafi Khardalian Senior System Administrator Pennyweb, Inc.

2 1

← Newer
1
2
3
4
5
6
7
8
...
26
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters May 2001