toasters

toasters@lists.teaparty.net

1 participants
13518 discussions

RE: cifs terminate username?
by Jain, Ankur 06 Jun '07

06 Jun '07

David, The "ONTAPI" referred in the mail below is actually Manage ONTAP SDK product. Manage ONTAP SDK provide set of interfaces for developing third party integrations to manage NetApp storage. The SDK provide interfaces in C, Java and Perl languages. SDK can manage NetApp storage running ONTAP 6.3 and above. Manage ONTAP SDK is currently used by numerous partners and customers to tightly integrate into their own tools and applications with NetApp storage systems. The SDK can be downloaded from SDK portal( http://now.netapp.com/cgi-bin/manageontap ). The access to this portal is restricted. NetApp customers can request access to SDK through their NetApp sales representative. Thanks, Ankur -----Original Message----- From: David Lee [mailto:t.d.lee@durham.ac.uk] Sent: Tuesday, June 05, 2007 7:38 PM To: Paul Rosham Cc: toasters(a)mathworks.com Subject: RE: cifs terminate username? On Tue, 5 Jun 2007, Paul Rosham wrote: > Perhaps you could try and swap your scripting to the XML-based ONTAPI, > which would allow much better machine parseability? That sounds interesting. Until your email I had never heard of 'ONTAPI'. A quick Google search showed a few hits, but no obvious authoritative source of either specification or tutorial. How long has it been in ONTAP? (We are 7.0.5) How does one start, from zero, to investigate this thing called 'ONTAPI'? Documentation? Examples? Support software/modules for 'perl' (or shell, or similar)? Thanks in advance. -- : David Lee I.T. Service : : Senior Systems Programmer Computer Centre : : UNIX Team Leader Durham University : : South Road : : http://www.dur.ac.uk/t.d.lee/ Durham DH1 3LE : : Phone: +44 191 334 2752 U.K. :

1 0

cifs terminate username?
by David Lee 05 Jun '07

05 Jun '07

(OnTap 7.0.5) From time to time our admin procedures mean that we have to move someone's home directory from one qtree to another. (I'd rather this didn't happen, but it's one of those things we have to live with for the time being...) Inevitably, on some occasions a CIFS session for such a user is active at the time of the move. So if, the morning after the move, they simply pick up from where they left off the previous evening, their new data is still going into their old area. (This probably happens one every week or two on average.) What we're looking for is something like: cifs terminate username that we could employ to kill off the CIFS session(s) for a given username. (Such killing of potentially active (usually dormant) sessions is (while not ideal) nevertheless usually far preferable to picking up the pieces of two diverging home directories in the following days.) But there doesn't seem to be any such "cifs terminate username". There is a "cifs sessions username", whose output is intended for humans, not machines, so is not easily amenable to reliable machine-parsing. So it is not really suitable for then reliably deriving workstation names for OnTap's existing "cifs terminate workstation" command. Any thoughts on how to achieve "cifs terminate username" functionality? By the way, the environment for driving this is a set of UNIX scripts which can do operations of the form: rsh filer cifs ... -- : David Lee I.T. Service : : Senior Systems Programmer Computer Centre : : UNIX Team Leader Durham University : : South Road : : http://www.dur.ac.uk/t.d.lee/ Durham DH1 3LE : : Phone: +44 191 334 2752 U.K. :

3 3

RE: cifs terminate username?
by David Lee 05 Jun '07

05 Jun '07

On Tue, 5 Jun 2007, Paul Rosham wrote: > Perhaps you could try and swap your scripting to the XML-based ONTAPI, which > would allow much better machine parseability? That sounds interesting. Until your email I had never heard of 'ONTAPI'. A quick Google search showed a few hits, but no obvious authoritative source of either specification or tutorial. How long has it been in ONTAP? (We are 7.0.5) How does one start, from zero, to investigate this thing called 'ONTAPI'? Documentation? Examples? Support software/modules for 'perl' (or shell, or similar)? Thanks in advance. -- : David Lee I.T. Service : : Senior Systems Programmer Computer Centre : : UNIX Team Leader Durham University : : South Road : : http://www.dur.ac.uk/t.d.lee/ Durham DH1 3LE : : Phone: +44 191 334 2752 U.K. :

2 1

Clustered heads with SATA disk shelves?
by Tom Samplonius 04 Jun '07

04 Jun '07

One item that I've researching that is a bit unclear, is that, can I can connect a SATA disk shelf (DS14mk2 AT) to a clustered set of FAS960 heads? I think there is no problem. I assume that you must use the AT-FCx modules, as they are dual-looped. I suspect that the the AT-FCx is mandatory for clustering (as opposed to the AT-FC) , but NOW seems a bit vague, at least according to what I was able to find. Also, I've heard rumors that connecting SATA shelves to a cluster is not really reliable. I was a bit taken a back, as NetApp as the SATA shelves have been out for a while. Is this really the case? Tom

7 9

Dual-domain CIFS
by Nils Vogels 30 May '07

30 May '07

Hi, In a cluster, since there are two heads, it is possible to make each head member of a different domain for CIFS. Will this cause problems during failover? Greets, Nils -- Simple guidelines to happiness: Work like you don't need the money, Love like your heart has never been broken and Dance like no one can see you.

2 1

R200 AT-FC2/FCX
by Carl Howell 30 May '07

30 May '07

We currently have an R200 with six AT-FC2 shelves. We want to add three AT-FCX shelves. Currently, we only have two loops and the AT-FC2's are spread across both of them. Should I move all of the AT-FC2's onto one loop and put the AT-FCX shelves on the other loop or just add the AT-FCX's to the current loops? Thanks, --Carl

5 4

Re: ndmpcopy for Linux
by Sandeep Cariapa 27 May '07

27 May '07

Hi, I was able to build it on RHEL4, running kernel 2.6.9-42.EL. Grab the bits from here: http://www.ndmp.org/download/ndmpcopy The mods I made to get it to build were only in 2 files: ./src/Makefile 37c37 < LIBS = -L$(LIBDIR) -lndmp -lsocket -lnsl --- > #LIBS = -L$(LIBDIR) -lndmp -lsocket -lnsl 39c39 < #LIBS = -L$(LIBDIR) -lndmp --- > LIBS = -L$(LIBDIR) -lndmp ./lib/ndmp_common.h 40c40 < #define XDR_AND_SIZE(func) (bool_t(*)(XDR*, ...))xdr_##func##,sizeof(func) --- > #define XDR_AND_SIZE(func) (bool_t(*)(XDR*, ...))xdr_##func ,sizeof(func) I did not run any exhaustive tests after I built it, but hopefully this will at least get you started. Cheers, Sandeep Cariapa Sales Engineer, IBRIX http://www.ibrix.com =============================================== Augie Schwer wrote: Has anyone built ndmpcopy for Linux recently? The Solaris version doesn't seem to build anymore and the ndmp.org web site does not look very active anymore.

2 1

Max number of LUNs
by Robert Lazzurs 27 May '07

27 May '07

Hello, In relation to the recent threads on max inodes and max dir size I was wondering if anyone has any experience with the limit on the number of LUNs in a filer? Kind regards,

2 6

Re: Max number of LUNs
by Robert Lazzurs 26 May '07

26 May '07

I presume this is one of the features coming out in the new 10.1 release? On 26/05/07, Paul Rosham <paulrosham(a)yahoo.com.au> wrote: > Yes, GX does not support LUNs yet > > -----Original Message----- > From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On > Behalf Of Robert Lazzurs > Sent: Saturday, 26 May 2007 11:31 AM > To: toasters(a)mathworks.com > Subject: Re: Max number of LUNs > > On 23/05/07, Fox, Adam <Adam.Fox(a)netapp.com> wrote: > > I don't think so, although you may want to confirm with tech support. > > I think once you clone a lun or create a snapshot-backed lun, that > > would count against the total. > > But just having the lun in a shapshot > > would not count. But I could be wrong here as this is not really one > > of my areas of expertise. > > Do you know if this is still a limitation in OnTAP GX as I cannot seem to > find the information about this, or does anyone running GX know what the LUN > limitations are? > > Take care, > >

1 0

RE: SnapDrive install with existing iSCSI LUNS
by Julio Calderon 24 May '07

24 May '07

If you want to test your procedure before you actually do it!! I would create a dummy lun and mount the same way dump some data in it, make sure you format the same type of file system.... Remove and try sd.... That wil raise your comfort level... My personal best practice is to try it, back it up, test restore and then do your live work.... Hope that helps Have fun Julio calderon -----Original Message----- From: Toasters [mailto:toasters@aegonusa.com] Sent: Thu May 24 12:59:08 2007 To: Nils Vogels; toasters(a)mathworks.com Subject: RE: SnapDrive install with existing iSCSI LUNS Just unmap them from the filer , install SD and then remap them with it. -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Nils Vogels Sent: Thursday, May 24, 2007 1:53 PM To: toasters(a)mathworks.com Subject: Re: SnapDrive install with existing iSCSI LUNS Hi Sean, On 5/24/07, Sean McWherter <sean(a)marinedepot.com> wrote: > > > What is the proper way to install snapdrive on a machine that already > has LUNs mapped with the iSCSI initiator? Should I disconnect the LUN > first? I did read one post on the NOW forums where a guy got hosed by > disconnecting the LUN and then installing SnapDrive after someone told > him that's what to do. He was never able to mount them back. I'm > trying to dot all my I's and cross all my T's before I dive in here. I'd say disconnecting the LUN, removing all traces in the iSCSI initiator, then installing SnapDrive, and then using SnapDrive to 'connect' to the LUN is the right way. If I were you, I'd test this using an ONTAP simulator, though ;) Grtz, Nils -- Simple guidelines to happiness: Work like you don't need the money, Love like your heart has never been broken and Dance like no one can see you.

1 0

SnapDrive install with existing iSCSI LUNS
by Sean McWherter 24 May '07

24 May '07

I need guidance in setting up SnapDrive on our Windows 2003 Enterprise x64 servers. We have two of these servers clustered hosting our SQL 2000 servers. My question is this: What is the proper way to install snapdrive on a machine that already has LUNs mapped with the iSCSI initiator? Should I disconnect the LUN first? I did read one post on the NOW forums where a guy got hosed by disconnecting the LUN and then installing SnapDrive after someone told him that's what to do. He was never able to mount them back. I'm trying to dot all my I's and cross all my T's before I dive in here.

3 2

rysnc and filers
by tmac 24 May '07

24 May '07

Okay, I am trying to copy data from a DOT-7 to a DOT-GX. I start off with: mount filer7:/vol/src /src mount filerg:/ /gx I then do (from RHEL4U2): cp -a /src/dir1 /gx I then do: rsyn -avH --delete /src/dir1/ /gx/dir1/ There is nothing changing on the source, yet it transfers files anyway. Anyone have a clue as to why? The destination get some files marked with current time, yet others are marked with an older time. -- --tmac RedHat Certified Engineer #804006984323821 Principal Consultant, RABA Technologies 240-373-3926 (office) 301-688-4705 (Lab) 214-279-3926 (eFAX)

3 2

Re: rysnc and filers
by Jerry 24 May '07

24 May '07

Or rsync --whole-file the first time, or ndmpcopy. ----- Original Message ---- From: Nils Vogels <bacardicoke(a)gmail.com> To: tmac <tmacmd(a)gmail.com> Cc: toasters(a)mathworks.com Sent: Wednesday, May 23, 2007 3:59:21 PM Subject: Re: rysnc and filers Hi tmac, On 5/23/07, tmac <tmacmd(a)gmail.com> wrote: > Okay, I am trying to copy data from a DOT-7 to a DOT-GX. > ... > > I then do (from RHEL4U2): > > cp -a /src/dir1 /gx > > I then do: > > rsyn -avH --delete /src/dir1/ /gx/dir1/ > > There is nothing changing on the source, yet it transfers files anyway. > > Anyone have a clue as to why? cp doesnt take all metadata. Try tar/cpio! # cd /source # tar cf - * | ( cd /target ; tar xf - ) -- Simple guidelines to happiness: Work like you don't need the money, Love like your heart has never been broken and Dance like no one can see you. ____________________________________________________________________________________Need a vacation? Get great deals to amazing places on Yahoo! Travel. http://travel.yahoo.com/

2 1

A-SIS (Netapp DeDupe)
by Glenn Dekhayser 23 May '07

23 May '07

Doing some reading on the A-SIS feature. Will this only De-Dupe WITHIN a flexvol, or leverage 4kb block commonalities within an aggregate? I would think the savings would be dramatic across flexvols, especially in Boot-SAN environments. Glenn (the other one)

2 1

running out of inodes problem
by Magnus Swenson 22 May '07

22 May '07

Hello Toasters, Just wanted to do a quick check, what the standard practise is when running out of inodes on a volume. I have several flex volumes in one aggregate. One of the volumes only at 80% full ran out of inodes. df -i will show number of inodes used and inodes free. This is a 100G volume with 3458831 inodes. According to now.netapp.com, there are two solutions, increase inodes with the 'maxfiles' command, or add more disk space to the volume. Has anybody had experience with this and which way did you go?

7 9

lun type: windows_gpt
by Buerger, Andreas 22 May '07

22 May '07

Hi, I just created a lun via lun setup on our 6070 running 7.2.2. I mentioned a lun type called windows_gpt, does anybody know what this type is used for? Is it a new feature in 7.2 ? Regards ----------------------------------------- Andreas Buerger Wincor Nixdorf Int. GmbH WN CIO4 Computing & Storage Heinz-Nixdorf-Ring 1 33106 Paderborn Phone: +49 5251 693 4215 Fax: +49 5251 693 4200 Mail: andreas.buerger(a)wincor-nixdorf.com -----------------------------------------

3 2

ndmpcopy for Linux
by Augie Schwer 22 May '07

22 May '07

Has anyone built ndmpcopy for Linux recently? The Solaris version doesn't seem to build anymore and the ndmp.org web site does not look very active anymore. -- Augie Schwer - Augie(a)Schwer.us - http://schwer.us Key fingerprint = 9815 AE19 AFD1 1FE7 5DEE 2AC3 CB99 2784 27B0 C072

2 1

Re: Clustered heads with SATA disk shelves?
by Wilkinson, Brent 21 May '07

21 May '07

Not sure but I believe you do not want to mix fc and sata loops. Could be wrong though. -----Original Message----- From: owner-toasters(a)mathworks.com <owner-toasters(a)mathworks.com> To: Tom Samplonius <tom(a)samplonius.org> CC: toasters <toasters(a)mathworks.com> Sent: Fri May 18 19:44:32 2007 Subject: Re: Clustered heads with SATA disk shelves? clustered sata shelves work just fine. I have several terabytes attached to a 3050 cluster with out any problems. AT-FCX is required for that to work. I'd also recommend being on the latest firmware on the AT-FCX. -Blake On 5/18/07, Tom Samplonius <tom(a)samplonius.org> wrote: > > One item that I've researching that is a bit unclear, is that, can I can connect a SATA disk shelf (DS14mk2 AT) to a clustered set of FAS960 heads? > > I think there is no problem. I assume that you must use the AT-FCx modules, as they are dual-looped. I suspect that the the AT-FCx is mandatory for clustering (as opposed to the AT-FC) , but NOW seems a bit vague, at least according to what I was able to find. > > Also, I've heard rumors that connecting SATA shelves to a cluster is not really reliable. I was a bit taken a back, as NetApp as the SATA shelves have been out for a while. Is this really the case? > > > Tom >

3 2

Symantec AV not releasing pBlks
by Horton, Jamie 18 May '07

18 May '07

Anyone ever had the below issue?: I'm running 2 Win2k3 servers as my Symantec Scan Engine Servers (running Symantec Anti-Virus for Network Attached Storage 4.3). Both act as primary vscan scanners (for my R200 and FAS960 filers). Every once in awhile one or both of my filers stop servicing CIFS requests, because they run out of pBlks (control blocks). After several outages this week, I have learned that in order for everything to run clean, I have to do the following in sequence: 1. vscan off (on both filers) 2. Stop scan engine service on both win2k3 servers 3. vscan on (on both filers) 4. Start scan engine service on both win2k3 servers If I have everything on, and I stop and start vscan only on the filers (without recycling services on the win2k3 servers), I immediately start losing pBlks until they are all exhausted, then nobody can access any shares. NetApp suggested I upgrade to 7.2.2 since this version deals with pBlks better. I upgraded last night, and since I didn't recycle scan engine services on the win2k3 servers, I ran out of pBlks early this morning. Once I executed the sequence above, all was well. NOTE: we are continuing to add more users to our environment (user home drives are on the FAS960), which obviously add more to the scanning load, but that doesn't account for it happening on my R200 (software shares and application shares). Any information would be greatly appreciated. Thanks, -Jamie

1 0

RE: Off Topic - brocade monitoring
by Alejandro Orellano 15 May '07

15 May '07

Hi Blake, You should use TX and RX FC Words. Note that these values HAS TO be multiplied by four to recieve number of bytes! OIDs: TX FC Words: 1.3.6.1.4.1.1588.2.1.1.1.6.2.1.11.(port+1) RX FC Words: 1.3.6.1.4.1.1588.2.1.1.1.6.2.1.12.(port+1) SNMP starts port count at 1, not at 0. So, for switch port 0 you should use 1.3.6.1.4.1.1588.2.1.1.1.6.2.1.11.1 and 1.3.6.1.4.1.1588.2.1.1.1.6.2.1.12.0 Finally, note that (RxWord + TxWord * 4) give the same results as PortPerfShow. More info on http://www.brocadeconnect.com (you need a user/pass). Regards, ./alex -----Mensaje original----- De: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] En nombre de Blake Golliher Enviado el: martes, 15 de mayo de 2007 2:24 Para: toasters(a)mathworks.com Asunto: Off Topic - brocade monitoring Anyone here know how to get MB/ Sec / Port metric out of a brocade 4100 using SNMP? I'll upgrade the firmware / OS if I have to, but I can't find anywhere in the MIB. Can I take the frames counter, and multiply it by 2116 (2112 byte frame + 4 byte header)? That'd only work if the frames were always fulll, which seems wrong. It's driving me crazy, and I was just hoping someone knew on this list. If there's a more brocade centric list, please let me know. Thanks, -Blake

1 0

Off Topic - brocade monitoring
by Blake Golliher 15 May '07

15 May '07

Anyone here know how to get MB/ Sec / Port metric out of a brocade 4100 using SNMP? I'll upgrade the firmware / OS if I have to, but I can't find anywhere in the MIB. Can I take the frames counter, and multiply it by 2116 (2112 byte frame + 4 byte header)? That'd only work if the frames were always fulll, which seems wrong. It's driving me crazy, and I was just hoping someone knew on this list. If there's a more brocade centric list, please let me know. Thanks, -Blake

1 0

offline ldap impacting cifs
by jay d 14 May '07

14 May '07

I have 2 clustered filers. One filer shares several vols via nfs. The security on these vols is set to unix. Some of them require access from windows in which case a cifs share has been created and ldap is used to map windows accounts to unix uids/gids. On the other filer, all vols are ntfs, there's no nfs (cifs only), and ldap is configured and enabled. We experienced an outage on all cifs connections on both filers when the one ldap server we were using went down. "net view" (from a windows client) would hang and explorer couldn't find the network path. I would expect the mapping to fail and windows users to be mapped to the default pcuser on the unix vols accessed via cifs. I would also expect the cifs only filer to have no issues continuing to serve cifs. Is my expectation off? I can resolve the issue by adding additional ldap hosts but I want to uncover why cifs only connections are dependent on ldap when my ad domain controllers are online. Thanks, jd

1 0

Problems with mixed mode
by Michael Fahy 14 May '07

14 May '07

We have an FAS960 running NFS and CIFS. There are entries in the cifs_homedir.cfg file so that Windows users can map drives to their Unix home directories. Windows users can successfully map drives to their unix home directories but unless their files and directories are configured with write permission for "everyone", (i.e. 666 permissions for files) the Windows users cannot write to their own files and directories. The filer is not successfully matching the Windows username with the corresponding Unix username (even though the names are the same). We are using OpenLDAP for authentication on the Unix side. We have been told that "Files and directories created from unix using nfs only get their permissions read properly by other unix systems, and the same with cifs by Windows systems" but I do not believe this. I'm sure we are just not configuring the filer correctly. Any suggestions?

4 3

RE: Duplicate nonboot vol0
by MCLACHLAN, Alan 11 May '07

11 May '07

I have had this occur a couple of times over the years. Note in Michael's case the two vol0's were on the same physical disks, so it wasn't a result of drive adds. If you search on NOW I think from memery there's a bug article in about a race condition where this can occur. I'm too busy to bother looking it up right now ;) Regards, Alan McLachlan Technical Analyst Information Management Services Department of Education, Training and the Arts (DETA) Floor 11 Education House, 30 Mary Street, Brisbane Fax: 07 3237 9695 Mob: 0429 655644 -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Dave Rubright Sent: Friday, May 11, 2007 10:41 AM To: Michael Homa; toasters(a)mathworks.com Subject: RE: Duplicate nonboot vol0 I'm curious as to how you ended up with 2 volumes called vol0. Meaning, you can't create a volume named vol0 if it already exists even if it's offline. My only guess would be that you added some drives that were part of another 2 disk volume called vol0 which ONTAP should have renamed to vol0(1) when inserted. Not sure what happened there but... So an option if you did add drives, you should know which drives you added and remove them. If you don't know which drives were added then a prior autosupport should tell you which drives were dedicated to the original vol0 and remove the ones not on the volume list. From here you should be able to rename original vol0 to another name (orig_vol) and bring online. Pop in the drives you removed and offline vol0, destroy it. Then you can rename "orig_vol" volume back to vol0. Dave Rubright Jr 4Base Technology, Inc > -----Original Message----- > From: owner-toasters(a)mathworks.com > [mailto:owner-toasters@mathworks.com] On Behalf Of Michael Homa > Sent: Thursday, May 10, 2007 3:59 PM > To: toasters(a)mathworks.com > Subject: Duplicate nonboot vol0 > > We have a backup F740, running 6.4.5, that serves as an > insurance to our production, primary, netapp system. And, > yes, before someone suggests it, we are in the process of > upgrading it to a 270 with a more recent version of ONTAP. > > Somehow a second vol0 (obviously traditional since the ONTAP > < 7) was created and I'm trying to figure out how to destroy it: > > Volume State Status Options > vol0 offline normal raidsize=14 > > Plex /vol0/plex0: online, normal, active > RAID group /vol0/plex0/rg0: normal > RAID group /vol0/plex0/rg1: normal > > vol0 offline normal raidsize=14 > > Plex /vol0/plex0: online, normal, active > RAID group /vol0/plex0/rg0: normal > RAID group /vol0/plex0/rg1: normal > > A disk in the duplicate vol0 is bad and that's how I > discovered the problem: > > Volume vol0 (restricted, normal) (block checksums) > Plex /vol0/plex0 (online, normal, active) > RAID group /vol0/plex0/rg0 (degraded) > > RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) > Phys (MB/blks) > --------- ------ --------------- ---- -------------- > -------------- > parity 7.48 7 3 0 FC:A 68000/139264000 > 69536/142410400 > data FAILED N/A 68000/139264000 > > I'm a bit unsure how to proceed. If I pull the disk, will the > system not grab one of the spares and rebuild on it? What I'd > like to happen is have the system rename the bad vol0 as > vol0(1) and then I can use the vol destroy command to get rid > of it. How do I do that? If I use the vol rename command, > there's the issue of identifying the "bad" vol0 and not the > one I want to retain. > > There was a ticket in the netapp knowledge base with a > procedure but that situation was for a duplicate boot vol0s > > I'm sure this is relatively simple yet I haven't come up with > a solution. > > Michael Homa > Operating Systems Support and Database Group Academic > Computing and Communication Center University of Illinois at Chicago > email: mhoma(a)uic.edu > >

1 0

Duplicate nonboot vol0
by Michael Homa 11 May '07

11 May '07

We have a backup F740, running 6.4.5, that serves as an insurance to our production, primary, netapp system. And, yes, before someone suggests it, we are in the process of upgrading it to a 270 with a more recent version of ONTAP. Somehow a second vol0 (obviously traditional since the ONTAP < 7) was created and I'm trying to figure out how to destroy it: Volume State Status Options vol0 offline normal raidsize=14 Plex /vol0/plex0: online, normal, active RAID group /vol0/plex0/rg0: normal RAID group /vol0/plex0/rg1: normal vol0 offline normal raidsize=14 Plex /vol0/plex0: online, normal, active RAID group /vol0/plex0/rg0: normal RAID group /vol0/plex0/rg1: normal A disk in the duplicate vol0 is bad and that's how I discovered the problem: Volume vol0 (restricted, normal) (block checksums) Plex /vol0/plex0 (online, normal, active) RAID group /vol0/plex0/rg0 (degraded) RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) Phys (MB/blks) --------- ------ --------------- ---- -------------- -------------- parity 7.48 7 3 0 FC:A 68000/139264000 69536/142410400 data FAILED N/A 68000/139264000 I'm a bit unsure how to proceed. If I pull the disk, will the system not grab one of the spares and rebuild on it? What I'd like to happen is have the system rename the bad vol0 as vol0(1) and then I can use the vol destroy command to get rid of it. How do I do that? If I use the vol rename command, there's the issue of identifying the "bad" vol0 and not the one I want to retain. There was a ticket in the netapp knowledge base with a procedure but that situation was for a duplicate boot vol0s I'm sure this is relatively simple yet I haven't come up with a solution. Michael Homa Operating Systems Support and Database Group Academic Computing and Communication Center University of Illinois at Chicago email: mhoma(a)uic.edu

2 1

Jump to page:

2025

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