toasters June 2006

toasters@lists.teaparty.net

159 participants
197 discussions

Snaphost dir showing in Automount
by Steven Lim 01 Jun '06

01 Jun '06

Hi Toasters, Any idea why when I run 'du -k .' command in a AUTOMOUNT mountpoint, the .snapshot dir shows & added in to the total space. This is misleading of the actually space in the directory. Likewise, running the same command in a MANUALLY MOUNT mountpoint doesn't show the .snapshot dir. Thanks for any help.

2 1

Undeliverable mail: RETURNED MAIL: DATA FORMAT ERROR
by MAILER-DAEMON＠cluster1.bresnan.net 01 Jun '06

01 Jun '06

Failed to deliver to 'owner-linux-vax(a)mithra.physics.montana.edu' SMTP module(domain mithra.physics.montana.edu) reports: host mithra.physics.montana.edu says: 550 <owner-linux-vax(a)mithra.physics.montana.edu>: Recipient address rejected: User unknown in local recipient table

1 0

Re: **SPAM** Snaphost dir showing in Automount
by Roger Stansfield 01 Jun '06

01 Jun '06

Steven, I found a similar thing when using find on a manually mounted and automounted mountpoint. (Solaris 2.6, not tried anything else but I will). However, find /home/filer/* (even as root) didn't traverse snapshots I also seemed to come across mount filer:/vol/vol1/mnt find /mnt/users traversed snapshots but mount filer:/vol/vol1/users find /mnt did not (or the other way round!). Something like that. Anyway in the end I concocted something that worked for me. Maybe some form of du -k * might help. Roger -- > Roger Stansfield, > Rowett Research Institute, Bucksburn, Aberdeen, AB21 9SB, Scotland, U.K. > Telephone: 01224 716677. Fax: 01224 715349. > mailto:R.Stansfield@rri.sari.ac.uk http://www.rowett.ac.uk/

1 0

RE: Managing Quotas from home-grown app
by Warkentin, Grant 01 Jun '06

01 Jun '06

FWIW... We use Data Fabric Manager to set our user quotas. DFM does not do tree quotas. I still have to edit the quota file and restart or resize (restart if it's a net new tree quota (OnTap 7.01)). I find that DFM is real handy for user quotas in our environment. We have several qtrees each with a default user quota. Only need to use DFM if I need to override the default quota for a user. By doing it this way I only need to manage the 200 users that need over rides. The rest get the default settings. DFM removes the need to do a Restart/Resize when adjusting user quotas. Currently not using group quotas (Quotaing only used for Windows users). Have no idea if DFM handles these or not. DFM is not the ultimate solution but it's price was right ($0.00). NOTICE - This communication is intended ONLY for the use of the person or entity named above and may contain information that is confidential or legally privileged. If you are not the intended recipient named above or a person responsible for delivering messages or communications to the intended recipient, YOU ARE HEREBY NOTIFIED that any use, distribution, or copying of this communication or any of the information contained in it is strictly prohibited. If you have received this communication in error, please notify us immediately by telephone and then destroy or delete this communication, or return it to us by mail if requested by us. The City of Calgary thanks you for your attention and cooperation. -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Stephen C. Losen Sent: Thursday, June 01, 2006 6:09 AM To: Ronan Mullally Cc: toasters(a)mathworks.com Subject: Re: Managing Quotas from home-grown app > > I've got a requirement to manage filer quotas from a home grown admin > application. The basic functionality would be: > > 1) read user quota for user X > 2) read group quota for group Y > 3) set user quota for user X > 4) set group quota for group Y > > 1) and 2) can be accomplished fairly easily via SNMP, however 3) > and 4) do not appear to be as straightforward. > > Having read the quotas chapter in the storage admin guide it appears > the only way to control quotas is via the /etc/quotas file. This means > that whenever a quota changes (which I'd expect to occur several times > a day) I need to re-write /etc/quotas and do a quota re-size or re-build. > > Is there no more elegant way of doing this - via SNMP perhaps? > You are stuck with editing the quotas file. You probably found out from the manual that netapp quotas are very complex, so that is why quotas are specified in a file. There is no way to set a single quota for a user or group on a netapp (unless you have a single volume with no qtrees in it). Each volume and qtree has its own user and group quotas, so one user can potentially have hundreds of different quotas. Fortunately the quotas file allows you to set default quotas, so you do not explicitly specify all of these quotas in practice. Steve Losen scl(a)virginia.edu phone: 434-924-0640 University of Virginia ITC Unix Support

1 0

RE: performance problems with filer and snapmirror
by Willeke, Jochen 01 Jun '06

01 Jun '06

As i did some tests with a Nearstore running Ontap 7.0.4, too. I can say that is it not related to the Ontap version. I can say after various tests is that all access to this system are slow as long as the snapmirror runs. And all snapmirror operations last very long. It takes about 1 minute to copy 1MB :-) After having a look at the network i saw not much traffic coming out of my filer. I will open a Call with Netapp as my knowledge is only limited ;-) Nevertheless, thanks a lot to all of you!! Regards Jochen -----Original Message----- From: Glenn Walker [mailto:ggwalker@mindspring.com] Sent: Thursday, June 01, 2006 2:38 PM To: philip.boyle(a)eircom.net; Willeke, Jochen Cc: toasters(a)mathworks.com Subject: RE: performance problems with filer and snapmirror Snapmirror will continue to work until Jochen moves to a . release (ie, 7.0 vs 7.1). He is currently running two releases in the 7.0 family. This is similar to using 6.4.1 vs 6.4.2 for comparison purposes. FWIW, QSM would not really care at all unless you had a file type that only existed on one release (ie, if you wanted to, QSM would work between 6.2 and 6.4 - but it wouldn't be able to move LUNs to the 6.2 system). Even though it should work, I wouldn't recommend it. As for the performance problem, I'd start with some questions: WHAT is slow (CIFS? NFS?) WHEN is it slow (Morning? Afternoon?) WHO is it slow for (all clients? Only some?) Then I'd communicate with NetApp Global Services (the support center in particular) to gather a perfstat and continue troubleshooting this issue. For example - if the CPU is never beyond 40% to 60%, the '-P flat' flag will probably not be of much use. FYI - Perfstat only automates command execution for things that already exist in ONTAP (many are in 'advanced' or 'diag' modes. If you wanted to, you could probably start with statit to see what is going on... but I'd definitely recommend speaking with NGS first, and they'll likely recommend perfstat as well. Glenn -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Philip Boyle Sent: Thursday, June 01, 2006 6:33 AM To: Willeke, Jochen Cc: toasters(a)mathworks.com Subject: Re: performance problems with filer and snapmirror Hi Jochen, I would have thought the snapmirror relationship would be now broken given the release differentiation? Should the R200 not require a minimum of ONTAP 7.0.4 in order to continue mirroring data from the source ( FAS3050 )? I found ONTAP 7.0.1R1 to be quite buggy and has numerous snapmirroring bugs fixed in ONTAP 7.0.4. In particular BURT 157883 which allowed me to oversubscribe my containing aggregate. http://now.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=157883 In order to debug your existing problem, I'd advise running 'perfstat' on FAS3050 and asking Netapp Support to review the output. perfstat.sh -f 'filername' -t 5 -i 3,0 -P flat >> 'logfile' Regards, Philip On Thursday 01 June 2006 09:17, Willeke, Jochen wrote: > Hello toasters, > > today we have a few problems with a FAS3050 (Ontap 7.0.4) which is very > slow, as the users tell me. When i do a sysstat i see the CPU always > between 40 and 60% which is not too much. > > I also had a look at the ocupied space in the aggregates and it should > be ok, too. According to this we have a snapmirror-relation to a > Nearstore R200 (Ontap 7.0.1R1) which is very slow. The network is not > the bottleneck as we have an own storagenet (1Gb Ethernet) for these > systems. > > I now wonder where is shall look for deeper analysis. I am a little > confused right now. Perhaps anybody can help with some kind of commands > different to sysstat :-) > > Best Regards > > Jochen -- ------------------ Philip Boyle Senior Systems Administrator Eircom.net philip.boyle(a)eircom.net tel: +353 1 701 0921 mob: +353 85 1744 354

1 0

RE: performance problems with filer and snapmirror
by Glenn Walker 01 Jun '06

01 Jun '06

Snapmirror will continue to work until Jochen moves to a . release (ie, 7.0 vs 7.1). He is currently running two releases in the 7.0 family. This is similar to using 6.4.1 vs 6.4.2 for comparison purposes. FWIW, QSM would not really care at all unless you had a file type that only existed on one release (ie, if you wanted to, QSM would work between 6.2 and 6.4 - but it wouldn't be able to move LUNs to the 6.2 system). Even though it should work, I wouldn't recommend it. As for the performance problem, I'd start with some questions: WHAT is slow (CIFS? NFS?) WHEN is it slow (Morning? Afternoon?) WHO is it slow for (all clients? Only some?) Then I'd communicate with NetApp Global Services (the support center in particular) to gather a perfstat and continue troubleshooting this issue. For example - if the CPU is never beyond 40% to 60%, the '-P flat' flag will probably not be of much use. FYI - Perfstat only automates command execution for things that already exist in ONTAP (many are in 'advanced' or 'diag' modes. If you wanted to, you could probably start with statit to see what is going on... but I'd definitely recommend speaking with NGS first, and they'll likely recommend perfstat as well. Glenn -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Philip Boyle Sent: Thursday, June 01, 2006 6:33 AM To: Willeke, Jochen Cc: toasters(a)mathworks.com Subject: Re: performance problems with filer and snapmirror Hi Jochen, I would have thought the snapmirror relationship would be now broken given the release differentiation? Should the R200 not require a minimum of ONTAP 7.0.4 in order to continue mirroring data from the source ( FAS3050 )? I found ONTAP 7.0.1R1 to be quite buggy and has numerous snapmirroring bugs fixed in ONTAP 7.0.4. In particular BURT 157883 which allowed me to oversubscribe my containing aggregate. http://now.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=157883 In order to debug your existing problem, I'd advise running 'perfstat' on FAS3050 and asking Netapp Support to review the output. perfstat.sh -f 'filername' -t 5 -i 3,0 -P flat >> 'logfile' Regards, Philip On Thursday 01 June 2006 09:17, Willeke, Jochen wrote: > Hello toasters, > > today we have a few problems with a FAS3050 (Ontap 7.0.4) which is very > slow, as the users tell me. When i do a sysstat i see the CPU always > between 40 and 60% which is not too much. > > I also had a look at the ocupied space in the aggregates and it should > be ok, too. According to this we have a snapmirror-relation to a > Nearstore R200 (Ontap 7.0.1R1) which is very slow. The network is not > the bottleneck as we have an own storagenet (1Gb Ethernet) for these > systems. > > I now wonder where is shall look for deeper analysis. I am a little > confused right now. Perhaps anybody can help with some kind of commands > different to sysstat :-) > > Best Regards > > Jochen -- ------------------ Philip Boyle Senior Systems Administrator Eircom.net philip.boyle(a)eircom.net tel: +353 1 701 0921 mob: +353 85 1744 354

1 0

performance problems with filer and snapmirror
by Willeke, Jochen 01 Jun '06

01 Jun '06

Hello toasters, today we have a few problems with a FAS3050 (Ontap 7.0.4) which is very slow, as the users tell me. When i do a sysstat i see the CPU always between 40 and 60% which is not too much. I also had a look at the ocupied space in the aggregates and it should be ok, too. According to this we have a snapmirror-relation to a Nearstore R200 (Ontap 7.0.1R1) which is very slow. The network is not the bottleneck as we have an own storagenet (1Gb Ethernet) for these systems. I now wonder where is shall look for deeper analysis. I am a little confused right now. Perhaps anybody can help with some kind of commands different to sysstat :-) Best Regards Jochen

2 1

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 June 2006