toasters December 2012

toasters@lists.teaparty.net

34 participants
16 discussions

Flexcache on 7.2.6.1
by Steven Kreuzer 07 Dec '12

07 Dec '12

I have an old 2050 running 7.2.6.1 which has been decommissioned in favor of new equipment running 8.0.2P4 I would like to redeploy the 2050 at a remote office as a flexcache volume of the new filer to improve the quality of life for the folks working remotely. I don't see any reason why this would be a bad idea, but is there anything I should be aware of when putting a flexcache volume on a filer running 7.2.6.1 (I am going to put the latest 7 code release on it before it gets deployed) in front of a filer running 8.0.2p4?

1 0

CIFS clients unable to access folders/ACL's until attached session is closed
by Ray Van Dolson 05 Dec '12

05 Dec '12

Got a strange one here... Have a folder (NTFS qtree w/ oplocks on) shared out via CIFS (also NFS, but expected to be minimal NFS access but still a variable). This folder contains files used for internal software builds/compiles and so several clients can be accessing it at the same time. CIFS clients in this case are typically Mac OSX. Ocassionally (lately, a couple of times a day), CIFS clients are unable to access portions of the tree. Even from Windows we get a permission denied error and can't even view ACL's on the affected folder. To solve the problem, we look for a CIFS client holding the location open and close it out. Then the folder becomes usable again. The CIFS client only has the folder open in "Read" mode. A couple thoughts from my side: - Maybe we should disable oplocks. Shouldn't be a lot (if any) NFS access to this share, but is a variable. - Our permissions are a bit wonky here. Share level permissions are in place to restrict most access to read-only, but file level permissions are set to full for "Everyone". Perhaps this is resulting in some sort of exclusive locking happening based on the more open nature of the file system permissions even if the share level allows read only? We'll probably try the oplocks thing for starters, but looking for any insights. Oh -- this is ONTAP 8.0.2P3 in 7-mode. Thanks, Ray

1 0

data sheets and spec?
by Mark Flint 05 Dec '12

05 Dec '12

Can anyone point me at a hardware matrix document, so I can work out how much disk I can support under a pair of heads? :) `Mark -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.

2 2

Re: nfsv4/parallel NFS bugs
by Scott Eno 03 Dec '12

03 Dec '12

Sounds like what they are seeing. The mount is obviously gone as writes are going to local disk, but the OS seems to insist the mount is still there. That's as I understand it, anyway. On Dec 3, 2012, at 9:45 AM, Mark Flint <mf1(a)sanger.ac.uk> wrote: > Just as an aside to the mount problem, are you having autofs issues underneath? i.e. the mount 'going away' > but still being in the mtab? > > > `Mark > > ext:6944 > > > > > On 3 Dec 2012, at 14:23, Scott Eno wrote: > >> >> Are there any NetApp bug reports you can point me to? >> >> We aren't using any of the "higher brain functions" of NFSv4 like file ownership and kerberos. UNIX admins started using it to fix a problem they saw on a volume shared by, up to, 17 hosts. Under NFSv3, when one host would write a file to the volume, other hosts wouldn't see the file. This problem goes away with NFSv4 mounts. Other problems, however, are appearing. A host will lose a mount to the NetApp and start writing to local disk. The RHEL OS, however, is adamant that the mount is still there. Very strange. >> >> They are also seeing "Bad Sequence ID" errors in the RHEL logs, but both NetApp and Red Hat claim this issue was resolved in earlier versions of their respective OS's. >> >> Now I'm wondering if 8.1.2 contains enough fixes for NFSv4 to cover our issues as well. >> >> >> >> On Dec 3, 2012, at 9:13 AM, tmac <tmacmd(a)gmail.com> wrote: >> >>> Should have mentioned the bugs that I have seen are fixed in 8.1.2 >>> >>> In my environment....performance related bugs. I was fine with a couple dozen hosts... >>> opening the floodgates to a couple hundred nfsv4 hosts broke. >>> Had to revert to nfsv3 (not a big deal, just revert the autofs table I was using). >>> >>> --tmac >>> >>> Tim McCarthy >>> Principal Consultant >>> >>> >>> >>> Clustered ONTAP Clustered ONTAP >>> NCDA ID: XK7R3GEKC1QQ2LVD RHCE5 805007643429572 NCSIE ID: C14QPHE21FR4YWD4 >>> Expires: 08 November 2014 Expires w/release of RHEL7 Expires: 08 November 2014 >>> >>> >>> >>> >>> On Mon, Dec 3, 2012 at 9:07 AM, Scott Eno <s.eno(a)me.com> wrote: >>> Hi, >>> >>> Your mentioned in the cluster thread of NFSv4 bugs caught my eye. Can you elaborate on that? We are moving RHEL 6.x servers to NFSv4 and a few odd issues have popped up (Bad Sequence ID errors, losing mounts). >>> >>> Running 8.1.1P1 across the board. >>> >>> On Dec 2, 2012, at 8:57 PM, tmac <tmacmd(a)gmail.com> wrote: >>> >>>> To the best if my knowledge, there is still no in place upgrade to clustered ontap. You must migrate to different heads. >>>> >>>> With that said, I have three different clusters... One large (4 fas 6280's) in my lab, one small(2 x fas 3240's) on my regular network and one more (2 x fas6080) used as backup or inter cluster snap mirror. >>>> >>>> Still 8.1.1p1d1 >>>> >>>> All I can say right now is watch out for a few nfsv4/parallel NFS bugs. Other than that... It had been real nice. >>>> >>>> On Dec 2, 2012 7:57 PM, "Peter D. Gray" <pdg(a)uow.edu.au> wrote: >>>> Hi people, >>>> >>>> I am contemplating going to cluster mode on our >>>> 3170A and was wondering what peoples experience has been? >>>> >>>> Any problems or gotchas? >>>> >>>> We are on 8.1.2. >>>> >>>> Regards, >>>> pdg >>>> >>>> _______________________________________________ >>>> Toasters mailing list >>>> Toasters(a)teaparty.net >>>> http://www.teaparty.net/mailman/listinfo/toasters >>>> _______________________________________________ >>>> Toasters mailing list >>>> Toasters(a)teaparty.net >>>> http://www.teaparty.net/mailman/listinfo/toasters >>> >>> >> >> _______________________________________________ >> Toasters mailing list >> Toasters(a)teaparty.net >> http://www.teaparty.net/mailman/listinfo/toasters > > > -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.

3 4

Re: nfsv4/parallel NFS bugs
by Scott Eno 03 Dec '12

03 Dec '12

This version? Linux automount version 5.0.5-54.el6 Directories: config dir: /etc/sysconfig maps dir: /etc modules dir: /usr/lib64/autofs Compile options: DISABLE_MOUNT_LOCKING ENABLE_IGNORE_BUSY_MOUNTS WITH_HESIOD WITH_LDAP WITH_SASL LIBXML2_WORKAROUND WITH_LIBTIRPC On Dec 3, 2012, at 10:03 AM, Mark Flint <mf1(a)sanger.ac.uk> wrote: > We had to upgrade to autofs5 to get round that issue, but I haven't tried V4 yet <gulp> :) > > > `Mark > > ext:6944 > > > > > On 3 Dec 2012, at 15:01, Scott Eno wrote: > >> >> Sounds like what they are seeing. The mount is obviously gone as writes are going to local disk, but the OS seems to insist the mount is still there. >> >> That's as I understand it, anyway. >> >> >> On Dec 3, 2012, at 9:45 AM, Mark Flint <mf1(a)sanger.ac.uk> wrote: >> >>> Just as an aside to the mount problem, are you having autofs issues underneath? i.e. the mount 'going away' >>> but still being in the mtab? >>> >>> >>> `Mark >>> >>> ext:6944 >>> >>> >>> >>> >>> On 3 Dec 2012, at 14:23, Scott Eno wrote: >>> >>>> >>>> Are there any NetApp bug reports you can point me to? >>>> >>>> We aren't using any of the "higher brain functions" of NFSv4 like file ownership and kerberos. UNIX admins started using it to fix a problem they saw on a volume shared by, up to, 17 hosts. Under NFSv3, when one host would write a file to the volume, other hosts wouldn't see the file. This problem goes away with NFSv4 mounts. Other problems, however, are appearing. A host will lose a mount to the NetApp and start writing to local disk. The RHEL OS, however, is adamant that the mount is still there. Very strange. >>>> >>>> They are also seeing "Bad Sequence ID" errors in the RHEL logs, but both NetApp and Red Hat claim this issue was resolved in earlier versions of their respective OS's. >>>> >>>> Now I'm wondering if 8.1.2 contains enough fixes for NFSv4 to cover our issues as well. >>>> >>>> >>>> >>>> On Dec 3, 2012, at 9:13 AM, tmac <tmacmd(a)gmail.com> wrote: >>>> >>>>> Should have mentioned the bugs that I have seen are fixed in 8.1.2 >>>>> >>>>> In my environment....performance related bugs. I was fine with a couple dozen hosts... >>>>> opening the floodgates to a couple hundred nfsv4 hosts broke. >>>>> Had to revert to nfsv3 (not a big deal, just revert the autofs table I was using). >>>>> >>>>> --tmac >>>>> >>>>> Tim McCarthy >>>>> Principal Consultant >>>>> >>>>> >>>>> >>>>> Clustered ONTAP Clustered ONTAP >>>>> NCDA ID: XK7R3GEKC1QQ2LVD RHCE5 805007643429572 NCSIE ID: C14QPHE21FR4YWD4 >>>>> Expires: 08 November 2014 Expires w/release of RHEL7 Expires: 08 November 2014 >>>>> >>>>> >>>>> >>>>> >>>>> On Mon, Dec 3, 2012 at 9:07 AM, Scott Eno <s.eno(a)me.com> wrote: >>>>> Hi, >>>>> >>>>> Your mentioned in the cluster thread of NFSv4 bugs caught my eye. Can you elaborate on that? We are moving RHEL 6.x servers to NFSv4 and a few odd issues have popped up (Bad Sequence ID errors, losing mounts). >>>>> >>>>> Running 8.1.1P1 across the board. >>>>> >>>>> On Dec 2, 2012, at 8:57 PM, tmac <tmacmd(a)gmail.com> wrote: >>>>> >>>>>> To the best if my knowledge, there is still no in place upgrade to clustered ontap. You must migrate to different heads. >>>>>> >>>>>> With that said, I have three different clusters... One large (4 fas 6280's) in my lab, one small(2 x fas 3240's) on my regular network and one more (2 x fas6080) used as backup or inter cluster snap mirror. >>>>>> >>>>>> Still 8.1.1p1d1 >>>>>> >>>>>> All I can say right now is watch out for a few nfsv4/parallel NFS bugs. Other than that... It had been real nice. >>>>>> >>>>>> On Dec 2, 2012 7:57 PM, "Peter D. Gray" <pdg(a)uow.edu.au> wrote: >>>>>> Hi people, >>>>>> >>>>>> I am contemplating going to cluster mode on our >>>>>> 3170A and was wondering what peoples experience has been? >>>>>> >>>>>> Any problems or gotchas? >>>>>> >>>>>> We are on 8.1.2. >>>>>> >>>>>> Regards, >>>>>> pdg >>>>>> >>>>>> _______________________________________________ >>>>>> Toasters mailing list >>>>>> Toasters(a)teaparty.net >>>>>> http://www.teaparty.net/mailman/listinfo/toasters >>>>>> _______________________________________________ >>>>>> Toasters mailing list >>>>>> Toasters(a)teaparty.net >>>>>> http://www.teaparty.net/mailman/listinfo/toasters >>>>> >>>>> >>>> >>>> _______________________________________________ >>>> Toasters mailing list >>>> Toasters(a)teaparty.net >>>> http://www.teaparty.net/mailman/listinfo/toasters >>> >>> >>> -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE. >> > > > -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.

1 0

On cluster mode
by Peter D. Gray 03 Dec '12

03 Dec '12

Hi people, I am contemplating going to cluster mode on our 3170A and was wondering what peoples experience has been? Any problems or gotchas? We are on 8.1.2. Regards, pdg

8 14

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 December 2012