toasters

toasters@lists.teaparty.net

13520 discussions

HP-UX R11 L2000 / F740 NFS Performance
by Dave Mills 29 Mar '00

29 Mar '00

Has anyone tried maximizing file loading performance in the above environment. I will be installing the F740 with 18 - 36GB drives soon and want to solve the problem before it exists. We will be loading aobut 200GB on each of 2 consecutive weekends. Subsequent daily processing will be very heavy batch sequential. Any comments appreciated. Dave Mills Cypress Systems, LTD

1 0

Re: Server capacity and performance
by Brian Hostetter 29 Mar '00

29 Mar '00

2-3K usually. The total number of OPS has not increased very much, it seems like writes are taking forever. I was thinking it may have soemthing to do with WAFL looking for the next open block and when the filer is pretty full it is hard to find. --Bian >>> Mike Horwath <drechsau(a)tiny.net> 03/29 9:32 AM >>> On Wed, Mar 29, 2000 at 08:46:14AM -0600, Brian Hostetter wrote: > Could my performance problems be related to space, or are they > probably just sheer volume of ops problems. What are the numbers for the # of ops per second? -- Mike Horwath Admin & Manager @ VISI.com WORK: drechsau(a)visi.com IRC: Drechsau http://www.visi.com/ HOME: drechsau(a)geeks.org The only Minnesota ISP with public statistics: http://noc.visi.com/ Garbage In -- Gospel Out. - berkeley fortune(6)

2 1

rc registry registry rc
by Jenkinson, John P (SAIC) 29 Mar '00

29 Mar '00

we get these messages on reboot of one netapp ** Option wafl.default_unix_user is being set to "pcuser" in /etc/rc, ** and this conflicts with a value - "root" - loaded from the registry. ** Commands in /etc/rc always override the registry at boot time, ** so the value of wafl.default_unix_user is now "pcuser". Fri Mar 24 16:15:10 AKS [rc]: Option wafl.default_unix_user with value "pcuser" in /etc/rc is overri from the registry. ** Option wafl.default_unix_user is being set to "root" in /etc/rc, ** and this conflicts with a value - "pcuser" - loaded from the registry. ** Commands in /etc/rc always override the registry at boot time, ** so the value of wafl.default_unix_user is now "root". the option line is not in the /etc/rc file anyone know what causes such schizo behaviour? wher is a netapp's registry? ======================== John Jenkinson BP Exploration (Alaska)

1 0

Re: Creating a RAID
by July at Zerowait 29 Mar '00

29 Mar '00

We're using 5.3.4R3. I'll be interested in hearing about your cautionary note regarding the usage of full disks or not. I'm pretty sure it's okay, though, because striping requires a parity drive and each RAID has its own parity drive. I don't see how WAFL can stripe across several RAIDs and properly maintain parity. July John Denholm wrote: > > > >From July at Zerowait on Tue, 28 Mar 2000 15:52:57 EST: > > > > > > > >Hello: The filer in question is an inherited box, an F760 with two > > > >shelves of 9GB SCSI drives, one shelf of 9GB FC-AL drives all in one > > > >RAID and one VOL. The single VOL also contains the root, I think. I've > > > >heard that you are supposed to separate SCSI and FC-AL drives into > > > >separate RAID groups, but this is how the box came and we can't change > > > >it. > > > > > > > >Anyway, my task is to add two shelves of 36GB drives to this unit. I > > > >know I need to create a separate RAID group for the 36GB drives. In the > > > >man pages I see instructions for adding disks to an existing volume and > > > >how to create a volume. I don't see anything that tells me how to create > > > >a new RAID group. I must be missing something obvious. > > July, (cool name :) > > A cautionary word... > > Someone from Netapp could perhaps address this one. > I'm not sure you'll get the full 36G out of your drives! > > I know it always used to be the case that if you had different raid sets with > different size drives in the raid sets, as long as the set was all the same > size, it didn't matter, all disks were fully used. This was because WAFL used > to round-robin between raid sets for writes. > > I heard a whisper that this wasn't the case any more, that now it fully striped > across ALL raid sets. Can someone from Netapp jump in with an AWOOGAH or an > all clear? > > So, what version of ONTAP are you running there July? > > J > > --------------------------------------------------------------------------- > # John Denholm johnd(a)theplanet.net # > # Webcache & Filer Administrator, Planet Online +44 113 207 6357 # > --------------------------------------------------------------------------- > > Error 404: There is no spoon -- ---------------------------------------------------------------- July A. J. Linett, VP Tech Mktg 302.266.9408 july(a)zerowait.com Zerowait: Best Of Breed Internet Core Technologies Web Acceleration ** Data Storage ** Load Balancing www.zerowait.com*www.loadbalancing.net*www.nas-san.com -------------------------------------------------------------------

1 0

RE: Re: NIS and netgroup file
by johnj.sasso＠ps.ge.com 29 Mar '00

29 Mar '00

I fully agree. Having experience with administering Auspex servers as well as NetApps, as well as administering a large, complex NIS environment, it would be a pity to start adding more overhead to the NetApp outside of what it was originally intended for (i.e. fileservice). The Auspex was was a pain to manage because its lack of simplicity, and I didn't see the dedication to fileservice that NetApp provides. As of 5.3.4 (?), NetApp provides the nis.servers option, where you can list all the possible NIS servers to bind to. This has worked quite well in our environment, esp. in the face of a failure of an NIS server the filer was bound to. Adding more overhead to the filer, such as NIS slave service, will only complexify the NetApp more, and as Jeff alluded to, bring it more into the Auspex realm (IMHO, this would only hurt NetApp). One of the features that makes the NetApp great to work with is its simplicity w.r.t. administration; if this isn't for you, then go with an Auspex or Sun (and buy enough Tylenol to boot). --john -----Original Message----- From: Jeff Innis [mailto:figaro911@hotmail.com] Sent: Tuesday, March 28, 2000 6:13 PM To: sirbruce(a)ix.netcom.com Cc: abond(a)netapp.com; demarest(a)arraycomm.com; toasters(a)mathworks.com Subject: Fwd: Re: NIS and netgroup file Sure why not add NIS server, why stop there though lets add DNS too. Hey these things are so reliable, let's put a full blown kernel in there, and start running user apps as well. Anyone ever hear of a little box called Auspex??? I actually thought the concept of "dedicated" NFS appliance was a good one. Jeff > > > We are looking into ways of improving the NIS implementation to be more > > efficient. We are not planning on implementing the slave functionality >on > > the filer though. > >That's too bad, because it essentially makes a network less robust. >See, if you rely on NIS files so you don't have to copy files to the >filer locally, then lets say you have a power outage. The filer comes >up and then can't do anything because it's waiting to talk to the NIS >server (it's probably the DNS server too). The NIS server can't do >anything because it's waiting to mount the filer. Stalemate. There are >ways around this, like having your NIS/DNS server hold the data on a >local disk, but then you have a kink in your whole centralized network >storage scheme. Altenatively, you can copy everything to the filer, but >then you've defeated the whole purpose of NIS. > >Making filers slave servers is an ideal solution. You don't have to >allow other clients to bind to them, but this way they can get data >updated from a master and still have a local copy in the event of >network problems. > >Bruce > > ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

1 0

RE: NIS and netgroup file
by Nagaraj, Devi 29 Mar '00

29 Mar '00

Jeff/Bruce, Your input is very valuable. We will take into account your suggestion while we improve the NIS services on the filer for the next release. Jeff - instead of binding using broadcast, you can secify the NIS servers you want to bind to in the order of preference. This will make sure that the same NIS server is not bombarded by all the filers. -Devi -----Original Message----- From: Jeff Krueger [mailto:jkrueger@qualcomm.com] Sent: Tuesday, March 28, 2000 4:00 PM To: Bruce Sterling Woodcock Cc: Nagaraj, Devi; Bond, Andrew; Timothy Demarest; toasters(a)mathworks.com Subject: Re: NIS and netgroup file We are in full accord with Bruce's comments below. Although recent enhancements to the NIS client have made it speedier and more robust, loss of communications with a NIS slave has consistently been one of our largest causes for NetApp services outages. Additionally, binding several NetApps to the same slave places tremendous strain on that slave due to continual hammering by the filers. A lightweight slave implementation inside DOT would be a welcome feature. As Bruce mentioned, it need not provide maps to other machines, but only for its own purposes. This would reduce network traffic and strain on existing boxes, free up an external dependency, and likely improve the speed of authentication which may make the filer run faster overall. As a side-note, our old Auspex ran a NIS slave for the same basic reason although it was horrible for machines to bind to. In broadcast mode, we found that our NetApps *loved* binding to it among the multitude of other slaves on that particular subnet. =) -- 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 >From "Bruce Sterling Woodcock" on Tue, 28 Mar 2000 14:10:10 PST: > > >----- Original Message ----- >From: Nagaraj, Devi <devi.nagaraj(a)netapp.com> >To: 'Bruce Sterling Woodcock' <sirbruce(a)ix.netcom.com>; Bond, Andrew ><abond(a)netapp.com>; Timothy Demarest <demarest(a)arraycomm.com> >Cc: <toasters(a)mathworks.com> >Sent: Tuesday, March 28, 2000 9:15 AM >Subject: RE: NIS and netgroup file > > >> We are looking into ways of improving the NIS implementation to be more >> efficient. We are not planning on implementing the slave functionality on >> the filer though. > >That's too bad, because it essentially makes a network less robust. >See, if you rely on NIS files so you don't have to copy files to the >filer locally, then lets say you have a power outage. The filer comes >up and then can't do anything because it's waiting to talk to the NIS >server (it's probably the DNS server too). The NIS server can't do >anything because it's waiting to mount the filer. Stalemate. There are >ways around this, like having your NIS/DNS server hold the data on a >local disk, but then you have a kink in your whole centralized network >storage scheme. Altenatively, you can copy everything to the filer, but >then you've defeated the whole purpose of NIS. > >Making filers slave servers is an ideal solution. You don't have to >allow other clients to bind to them, but this way they can get data >updated from a master and still have a local copy in the event of >network problems. > >Bruce > >

1 0

RE: 3rd Party NFS Clients
by Jen Hamilton 29 Mar '00

29 Mar '00

For those of you using NFS Maestro to log onto NetApps from NT workstations, what authenticates the user? I get "Authorization Error" when logging in, unless I'm user "nobody" with no password. Thanks, Jen Hamilton

2 1

ndmpcopy problem (long)
by Deron Johnson 28 Mar '00

28 Mar '00

I had a need to transfer transfer a large amount of data between two filers, one running OnTAP 5.3.4 and one running OnTAP 5.3.4R2. I ran a level 0 ndmpcopy the day before the cutover, and then a level 1 the night of the cutover, in order to minimize downtime. However, after doing a compare of file sizes after the transfer, I found that some files which had been modified since the level 0 did not have the proper size on the destination. Here is an example. Toaster1 is the source, and toaster2 is the destination. I have added some comments in UPPER CASE. -------- # cd /toaster1/test # ls -l total 80 -rw-r--r-- 1 root other 37817 Jan 31 15:54 testfile SIZE OF ORIGINAL FILE IS 37817 BYTES. # rsh toaster1 snap create vol0 testndmp creating snapshot.................................... # ndmpcopy toaster1:/vol/vol0/.snapshot/testndmp/test toaster2:/vol/vol0/test -level 0 Password: Connecting to toaster1. Connecting to toaster2. toaster1: CONNECT: Connection established. toaster2: CONNECT: Connection established. toaster1: LOG: DUMP: toaster1: LOG: Using subtree dump toaster1: LOG: DUMP: toaster1: LOG: Date of this level 0 dump: Mon Jan 31 15:55:38 2000. toaster1: LOG: DUMP: toaster1: LOG: Date of last level 0 dump: the epoch. [NORMAL LOOKING DUMP/RESTORE OUTPUT] toaster1: LOG: DUMP: toaster1: LOG: DUMP IS DONE toaster1: HALT: The operation was successful! Waiting for toaster2 to halt too. toaster2: HALT: The operation was successful! The transfer is complete. Elapsed time: 0 hours, 0 minutes, 37 seconds. # rsh toaster1 snap delete testndmp deleting snapshot.................................... # cd /toaster2/test # ls -l total 68928 -rw-r--r-- 1 root other 37817 Jan 31 15:54 testfile -rw-rw-rw- 1 root root 35213312 Jan 31 15:57 restore_symboltable SIZE OF COPY IS 37817 BYTES, AS EXPECTED. # cd /toaster1/test # vi testfile # ls -l total 80 -rw-r--r-- 1 root other 37639 Jan 31 16:02 testfile I HAVE MODIFIED THE FILE, SO IT IS NOW SHORTER (37639 BYTES). # rsh -n toaster1 snap create vol0 testndmp creating snapshot.................................... # ndmpcopy toaster1:/vol/vol0/.snapshot/testndmp/test toaster2:/vol/vol0/test -level 1 Password: Connecting to toaster1. Connecting to toaster2. toaster1: CONNECT: Connection established. toaster2: CONNECT: Connection established. toaster1: LOG: DUMP: toaster1: LOG: Using subtree dump toaster1: LOG: DUMP: toaster1: LOG: Date of this level 1 dump: Mon Jan 31 16:03:47 2000. toaster1: LOG: DUMP: toaster1: LOG: Date of last level 0 dump: Mon Jan 31 15:55:38 2000. [NORMAL LOOKING DUMP/RESTORE OUTPUT] toaster1: LOG: DUMP: toaster1: LOG: DUMP IS DONE toaster1: HALT: The operation was successful! Waiting for toaster2 to halt too. toaster2: HALT: The operation was successful! The transfer is complete. Elapsed time: 0 hours, 0 minutes, 38 seconds. # cd /toaster2/test # ls -l total 68928 -rw-r--r-- 1 root other 37817 Jan 31 16:02 testfile -rw-rw-rw- 1 root root 35213312 Jan 31 16:05 restore_symboltable THE TIMESTAMP ON THE COPY HAS CHANGED, BUT IT IS STILL 37817 BYTES! # cd / # umount /toaster2 # mount toaster2:/vol/vol0 /toaster2 # cd /toaster2/test # ls -l total 68928 -rw-r--r-- 1 root other 37817 Jan 31 16:02 testfile -rw-rw-rw- 1 root root 35213312 Jan 31 16:05 restore_symboltable UNMOUNTED AND REMOUNTED TO MAKE SURE IT WASN'T A CACHING ISSUE. -------- Can anyone tell me what I might have done wrong? -- Deron Johnson djohnson(a)amgen.com

3 3

RE: NIS and netgroup file
by Nagaraj, Devi 28 Mar '00

28 Mar '00

Do you have NIS for netgroup lookup in your /etc/nsswitch.conf file? The group information ( not netgroup) can be obtained using the command ypgroup. -Devi -----Original Message----- From: Timothy Demarest [mailto:demarest@arraycomm.com] Sent: Tuesday, March 28, 2000 9:30 AM To: Bond, Andrew Cc: toasters(a)mathworks.com Subject: Re: NIS and netgroup file >>>>> "Andrew" == Bond, Andrew <abond(a)netapp.com> >>>>> wrote the following on Tue, 28 Mar 2000 00:53:07 -0800 Andrew> The filer does not copy the NIS maps into /etc. NIS group Andrew> information is cached, but otherwise the filer acts as a NIS Andrew> client only, and will query the NIS server it has bound to ("nis Andrew> info" will tell you which this is). Andrew: Thanks for the info. I can't seem to get the Filer to lookup netgroup information via NIS. If I manually place a system's name in the exports file, the system can then mount the volume. If I use an NIS netgroup name for which the system is a member, I cannot mount the volume. Is there a way to see what NIS information the Filer thinks it has cached? Tim

3 2

Re: NIS and netgroup file
by Timothy Demarest 28 Mar '00

28 Mar '00

>>>>> "Mark" == Mark Stoltzfus <stoltzfu(a)susq.com> >>>>> wrote the following on Tue, 28 Mar 2000 13:41:23 -0500 Mark> Do you use FQDNs in your netgroup map? If so, you're getting hit Mark> by bug #19922 (same as us :P ). We eventually threw our hands in Mark> the air and hacked our yp make script to just stick a copy of Mark> netgroup into our filers' /etcs. Sigh. Thanks, Mark. That seems to be the problem. I guess I'll hack the makefile to copy over the netgroup file. Tim

1 0

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