toasters June 2000

toasters@lists.teaparty.net

121 participants
153 discussions

RE: CIFS file locking problem
by mark.muhlestein＠netapp.com 21 Jun '00

21 Jun '00

The underlying problem may be due to a timing bug in the Microsoft client (MS has acknowledged the bug; their answer is "upgrade to Windows 2000") This is seen more on filers than NT server because we are so much faster doing opens. We have a workaround, which is to enforce a small delay before breaking the oplock on oplocked files that are re-opened. This is controlled by an option, "cifs.oplocks.opendelta". It defaults to 8 ms which works well for most clients, but it sounds like you might want to try increasing it. The "cifs stat" console command has a value "OpLkBkNoBreakAck" which gets incremented if this problem is seen. If you have non-zero values for this, try increasing the opendelta to 16 and see if things get better. As to the protocol issues, as Bruce said, we're more or less stuck with them because of Microsoft compatibility. If a session goes away when there is no request pending (i.e. we don't owe the client a response), the filer delays 10 minutes before terminating the session. This is to keep the session alive in the face of temporary network partitions, which we can't distinguish from just losing a client. Currently that 10 minute interval is not configurable on the filer, but it sounds like we might want to make it so. Mark Muhlestein -- mmm(a)netapp.com > -----Original Message----- > From: Dave Atkin [mailto:dla1@york.ac.uk] > Sent: Friday, June 16, 2000 3:39 AM > To: toasters(a)mathworks.com > Cc: AB Smith > Subject: CIFS file locking problem > > > We have a problem with file locking which occurs every few > months. We have > finally worked out at least part of what is happening. > > We have a Windows 95 system filestore exported as a read-only > CIFS share. > There are 1200 or so PCs which access this. > > Just occasionally when a user launches the Telnet client (Hummingbird) > it complains that it can't access its keyboard mapping file. Once this > starts to happen, it is the same for every PC. A unix user > running as root > can't read or do anything to this file either - it gets "permission > denied". > > Yesterday when this happened we used (rc_toggle_basic) lock_dump -f > to look at the locks, and we found that a PC appeared to have > taken out an > exclusive read lock on the file: > > ========00003a31:0006b48e > \95APPLIB\york.hts > state=GRANTED mode=Excl-denyA host=OFFICE1 > > When we found the PC called "office1", it was turned off. The > user said it > had hung up and she had switched it off. > > We managed to restore normal service by doing a > "cifs terminate -t 0 office1". > > This raises several issues: > > 1) If a CIFS client is switched off, is there any sort of > keepalive timer > on the Filer that closes down open files and locks? > > 2) Should a PC with read-only access to a filestore be allowed to lock > files for exclusive read? Should this perhaps be an option on > a qtree or > share? > > 3) This particular application obviously doesn't expect this > file to be > shared. But could any malicious user take out locks on random > important > files and freeze the whole network? > > Dave Atkin > -- > Dave Atkin, Head of Technical Services > Computing Service, University of York, YORK YO10 5DD > Phone: +44-1904-433804 (ddi) Fax: +44-1904-433740 > Email: D.Atkin(a)york.ac.uk >

2 1

Budtool "Migration"
by Jay Orr 20 Jun '00

20 Jun '00

Humm... Even though PDC has been claiming for the past 14+ months : http://www.pdc.com/BTMigration.html That they'll migrate current customers, I just heard from our rep that they aren't <really> offering a migration. Rather, a tradein to go to Legato Networker. For the record, the price of Legato minus the tradein offered is about the same (slight bit more) as orkstation would charge for a new installation. Haven't compared other backup solutions because when I originally contacted our rep back in February he thought it'd only be a few thousand. Granted, he had no clue either and no one there knew anything since Legato bought PDC, then decided to leave them to deal with budtool. So, for those of you who have to deal with budgets, be aware that you have 18 months before you're stuck with a backup system that no one will answer questions about. They say it's EOL'd Jan 2001, and they'll offer support for 12 months after that (they also said they'd migrate people too). Don't know about any of you, but I'm not happy.... ----------- Jay Orr Systems Administrator Fujitsu Nexion Inc. St. Louis, MO

1 0

Make /etc into a qtree?
by Jeff Krueger 20 Jun '00

20 Jun '00

Has anyone tried making /etc into a qtree? We'd like to do that and give it an infinite quota. That way the sum of our qtree usages will equal our df usage and we can track the size of /etc. I wonder if this could be done on a production filer assuming nobody is accesing /etc: adminhost% sudo rsh filer qtree create etc.new adminhost% sudo tarcp /etc /etc.new adminhost% sudo mv /etc /etc.old adminhost% sudo mv /etc.new /etc Seems like this could hose up syslog and possibly CIFS since it tinkers with its files there. Any ideas? Anyone tried this before? -- Jeff -- ---------------------------------------------------------------------------- Jeff Krueger E-Mail: jeff(a)qualcomm.com NetApp File Server Lead Phone: 858-651-6709 IT Engineering and Support Fax: 858-651-6627 QUALCOMM, Incorporated Web: www.qualcomm.com

5 4

SnapMirror reporting woes
by kevin graham 19 Jun '00

19 Jun '00

Perhaps I missed something, but does anyone know of any useful commands for determining snapmirror performance? I'm currently running 5.3.5R2. toaster1.iad> vol snapmirror status Source Dest Status toaster2.snv.xxxx:vol1 toaster1.iad.xxxx:vol2 Transferring (1% complete) toaster1.snv.xxxx:vol1 toaster1.iad:xxxx:vol1 Transferring (0% complete) This isn't exactly useful data, as they seem to almost always be at 0%, though I rarely ever see 'mirror postponed' messages, which inclines me to think that progress is being made in a timely manner. Updates are every 2 hours, and for two small-delta 60gv volumes, I figure it should be idle a fair deal of the time. I can watch recent snapshots show up on the snapmirror dest, but this is the only indication I have that its working at all. The main point of concern, is that Snapmirror dest filer has about 3300 miles between it and the source -- so we are much more susceptible to quirks in the public network than one would encounter in a lab environment or even corporate WAN. The best data I could find was from the snapmirror ops guide, which suggest looking at the snap list on the source filer -- SnapMirror generated Snapshots should have the name of the destination SnapMirror volume appended with a counter. By coordinating the snapmirror.conf file with the time stamps shown in the snapshot listing, you should be able to determine whether or not SnapMirror is generating Snapshots as scheduled ..this seems pitifully inadequate -- something akin to calculating your nfs ops/s by dividing nfsstat -s by uptime whenever you wanted to know. I've got syslog tweaked up to *.info, which hasn't given me much more than statd reports. Is there something better? Something that atleast shows some simple counters like: toasterN> vol snapmirror status -v Volume Source vol1 toaster1:vol1 Current status: idle Updates since reconfig: 94127 Failed updates: 5 Postponed updates: 17 Last successful update: Jun 16 15:12 (45m ago) 1251k, 15 minutes (1.39 KB/s) Average: 4912k, 22 minutes (3.70 KB/s) This would actually give something that snapmirror.conf's could be tuned with. Informative data to syslog would be good as well, use a local facility for it, and for those of us with >1hr replications could actually check progress via logs. Given the price tag on this option, I made the obviously foolish presumption that there would be sufficient tools to monitor progress and performance, especially in the light of its branding as a 'Disaster Recovery' solution.. Wishing I'd stuck to it and put the database on EMC w/ Oracle replication, ..kg..

3 2

restore problems
by Artur Shnayder 18 Jun '00

18 Jun '00

Hello all, I use dump utility for backup my filer on the tape device, that is attached to the filer. Here is the dump command: rsh netapp dump 0dsbfu 54000 52000 64 nrst0a /vol/vol0/home Now I need to restore single file: /vol/vol0/home/software/user/file. I understood from netapp guy, that it's inpossible. My question is how I can restore at least parent directory for this file /vol/vol0/home/software/user. Is it possible? If I run this command (it's the fifth record): netapp> restore xvfbs rst0a 64 5 /vol/vol0/home/software/user/file it ignores the last argument and tries to restore whole the record /vol/vol0/home. And if I add "D" options, like : netapp> restore xvfbsD rst0a 64 5 rst0a /vol/vol0/home/software/user/file /vol/vol0/home/software/user/tmp I get error message "/vol/vol0/home/software/user/tmp is not on tape". It should be very simple, but netapp support guy said me, that -s 5 and -b 64 options are eligible and advised to use the following format ( I guess it's from now.netapp.com): restore xvfD rst0a /vol/vol0/home/software/user/tmp /vol/vol0/home/software/user/file , that of course didn't work. Any help will be very appreciate. Thanks. ~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._. Artur Shnayder Voice: (609)655-5105 x12 System Manager Fax: (609)655-5114 Email: artur(a)cgen.com Compugen Inc URL: http://www.cgen.com ~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._.,-*~`^`~*-,._.

2 2

Re: Lost password
by Gustavo Chaves 17 Jun '00

17 Jun '00

Hi, I want to thank everyone who sent a message trying to help me with my problem. I also want to apologize for NOT HAVING READ your answers before. Yes. I just re-subscribed to the list to send the question and forgot that I had set up a filter long ago to automatically redirect the list's messages to a folder. :-( This was a particularly hard week and I didn't have the time to find out why I wasn't receiving the list's messages. Two days ago I powered down the filer and used a diskete boot to alter the password. On Monday I'll try some of your suggestions so that I don't fall in this trap in the future. Thanks, again. Gustavo.

1 0

CIFS file locking problem
by Dave Atkin 16 Jun '00

16 Jun '00

We have a problem with file locking which occurs every few months. We have finally worked out at least part of what is happening. We have a Windows 95 system filestore exported as a read-only CIFS share. There are 1200 or so PCs which access this. Just occasionally when a user launches the Telnet client (Hummingbird) it complains that it can't access its keyboard mapping file. Once this starts to happen, it is the same for every PC. A unix user running as root can't read or do anything to this file either - it gets "permission denied". Yesterday when this happened we used (rc_toggle_basic) lock_dump -f to look at the locks, and we found that a PC appeared to have taken out an exclusive read lock on the file: ========00003a31:0006b48e \95APPLIB\york.hts state=GRANTED mode=Excl-denyA host=OFFICE1 When we found the PC called "office1", it was turned off. The user said it had hung up and she had switched it off. We managed to restore normal service by doing a "cifs terminate -t 0 office1". This raises several issues: 1) If a CIFS client is switched off, is there any sort of keepalive timer on the Filer that closes down open files and locks? 2) Should a PC with read-only access to a filestore be allowed to lock files for exclusive read? Should this perhaps be an option on a qtree or share? 3) This particular application obviously doesn't expect this file to be shared. But could any malicious user take out locks on random important files and freeze the whole network? Dave Atkin -- Dave Atkin, Head of Technical Services Computing Service, University of York, YORK YO10 5DD Phone: +44-1904-433804 (ddi) Fax: +44-1904-433740 Email: D.Atkin(a)york.ac.uk

2 1

RE: Raid Groups and Volumes
by Deborah.Rousseau＠netapp.com 15 Jun '00

15 Jun '00

Here's the scoop on volumes, raid groups, parity drives, hot spares, qtrees, and root: Volume, Raid Groups, Parity, Hot Spares-- ----------------------------------------- A volume by definition will contain one or more raid groups, just like a tray of ice can contain one or more cubes of ice. A raid group can contain a maximum of 28 disks (one parity disk and 27 data disks). The default number of drives in a raid group is 14; that number is considered optimum for performance vs. fault tolerance. (There is data regarding MTTDL (Mean time to data loss) and other performance and mean time measures that prove 14 disks to be the optimum size for raid group.) (NetApp 202 teaches diagnostics and optimization of performance and troubleshooting.) You can setup your 14 - 36Gb drives into one volume. But it would be better to use 13 disks in the raid group and save one disk on shelf as your hot spare. In case of failure, your filer will immediately begin to reconstruct your data to your hot spare. If you have AutoSupport turned on and configured, Network Appliance will be notified by your filer (by email) that a disk has failed and a new disk will be sent to you automatically per our service agreement. Yes, hot spares and AutoSupport are great to have working for you.) Qtrees, security styles, oplocks, quotas, backup-- -------------------------------------------------- You can organize your one (13 drive) volume into qtrees. Qtrees are almost like separate file system partitions. Each qtree can be setup with its own set of properties. To be specific, each qtree can have its own: - security styles (UNIX, MIXED, or NTFS) - oplocks settings (ON or OFF) - quotas for limiting users, groups or limiting the size of the qtree itself - backup and restore schedules Snapshots are volume wide; you cannot take a snapshot of an individual qtree; however, all qtrees will be included in the volume's snapshot. If you setup 2 volumes, you will have a minimum of 2 raid groups and therefore use 2 of your 14 - 36Gb drives as parity drives. Smaller raid groups give you greater protection against a double-disk failure, but slightly less than optimum performance and less storage space. Also, if you reserved one disk as hot spare, it would serve both volumes. Hot spares are global to the filer and can be used in a disk failure of any volume when there are multiple volumes on the filer.) Root-- ------ Some like to setup root as a separate 2-disk volume (parity and data disks). Having root separate helps with backup and disaster recovery and perhaps there are other considerations I don't know about. If root is a separate 2-disk volume, of your 14 disks you would have 2 parity, 1 hot spare and 11 data disks. If you have one volume with one raid group, you will have 12 data disks. There are several other considerations about qtrees, quotas, backup, disaster recovery, snapshots and hot spares to be understood. I would highly recommend NetApp 101 and 202 for comprehensive, hands-on training on fine points such as these. You can find training schedules as well as information on NetApp Certification at http://www.netapp.com/education/#training. Also, a great place to look up technical answers is NetApp's NOW site: http://now.netapp.com/. Best regards, Deborah Deborah Rousseau, MCSE, MCT Technical Trainer Network Appliance, Inc. http://now.netapp.com/ -----Original Message----- From: Brian Hostetter [mailto:BrianH@dice.com] Sent: Wednesday, June 14, 2000 8:53 AM To: toasters(a)mathworks.com Subject: Raid Groups and Volumes Hello, I just got my two new 760s with 2 shelves of 36G drives each. I am going through the setup and have a question on Volumes and Raid Groups. Can I create two shelves into one big raid group in order to take advantage of all the spindles, and then create two volumes on that raid group? Or do I have to create 2 raid groups (thus using an extra parity drive) and have 1 volume living on 1 raid group. --Brian _______________________________________________ Dl-toasters mailing list Dl-toasters(a)netapp.com http://mailhost.corp.netapp.com/mailman/listinfo/dl-toasters

1 0

Removing vol0
by Brian Hostetter 15 Jun '00

15 Jun '00

I got two filers recently each configured a different way. They each contain two shelves. One had 2 drives used and the other marked spare, which was nice because I could config it any way I wanted. The other has all the drives setup into one 300G volume. Is it possible to pair a volume back down to just the two drives and create another volume? --Brian

2 1

RE: Raid Groups and Volumes
by jeffrey.soto＠netapp.com 15 Jun '00

15 Jun '00

Brian, You can have a volume with one or more RAID groups, but not a RAID group with multiple volumes. Disks added to a volume will automatically be added to the first existing raid group until that raid group reaches it's maximum size. After the maximum number of disks per the RAID group "raidsize" is reached, a new raidgroup is automatically created with a parity disk. However, you can create logical partitions on the volume using qtrees. Raid size is set either at volume creation by ; vol create volname [ -r raidsize ] example: filer > vol create vol1 -r 14 or on an existing volume by vol options volname optname optval example: filer > vol options vol0 raidsize 21 ( sets the size of the raid size on existing raid group and subsequent ones for this volume ) >From the manual; The default number of disks in a RAID group (including the parity disk) is 14 ( 2 shelves). A RAID group must contain at least two disks. The largest RAID group size you can create manually is 28 disks. The maximum RAID group size is 52 disks. Using smaller RAID groups reduces disk reconstruction time if a disk fails. Using smaller (therefore, more) RAID groups provides higher reliability by reducing the risk of data loss due to multiple-disk failure. Conversely, configuring with larger (therefore, fewer) RAID groups in a filer uses fewer disks for parity, leaving more disks available for data storage. Jeff Soto SE NYC NetworkAppliance, Inc mailto:jsoto@netapp.com Mobile : 1-516-635-0668 CorpVM : 1-408-822-8567 Pager : 1-877-563-8277 eFax: 1-419-831-5019 -----Original Message----- From: Brian Hostetter [mailto:BrianH@dice.com] Sent: Wednesday, June 14, 2000 11:53 AM To: toasters(a)mathworks.com Subject: Raid Groups and Volumes Importance: High Hello, I just got my two new 760s with 2 shelves of 36G drives each. I am going through the setup and have a question on Volumes and Raid Groups. Can I create two shelves into one big raid group in order to take advantage of all the spindles, and then create two volumes on that raid group? Or do I have to create 2 raid groups (thus using an extra parity drive) and have 1 volume living on 1 raid group. --Brian _______________________________________________ Dl-toasters mailing list Dl-toasters(a)netapp.com http://mailhost.corp.netapp.com/mailman/listinfo/dl-toasters

1 0

← Newer
1
...
4
5
6
7
8
9
10
...
16
Older →

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 June 2000