toasters September 2000

toasters@lists.teaparty.net

122 participants
161 discussions

CIFS style tape backups
by Clarke, John L. (CIS) 28 Sep '00

28 Sep '00

Hello, We have an F740 in a mixed NFS/SMB environment and have been trying to do CIFS style tape backups to preserver the ACL's. Unfortunately, HP Omniback II has been giving us fits and we need to look for other options. Are any of you doing CIFS style backups? What software are you using? How happy are you with that software, and the service provided by the vendor? Thanks John

2 1

RE: Low SMB throughput
by Mike Sphar 28 Sep '00

28 Sep '00

Have you run any network analyzers on the network? This is the kind of thing where a good sniffer (or a fluke, damn I love those) can be invaluable. "If it works, it must be a Fluke." -- Mike Sphar - Sr Systems Administrator - Engineering Support Services, BOFH, GWP, MCP, MCP+I, MCSE, BFD -----Original Message----- From: chris_good(a)webtop.com [mailto:chris_good@webtop.com] Sent: Thursday, September 28, 2000 6:56 AM To: toasters(a)mathworks.com Subject: Re: Low SMB throughput andrew-m.webster(a)db.com wrote: > Where is the delay? Is it that the inital connection to the Filer takes a long time? This would indicate that the Filer is having problems authenticating users via the domain controller. People log on OK, intermittently files just will not save, transfers take forever while nfs is still screaming. Netapp engineers have had a look at it and are as baffled as we are. >BTW: We never rely on autonegotiate (AUTO), but set each Filer interface to 100meg/Full Duplex. We force everything as a matter of course. -- Chris Good WebTop.com http://www.webtop.com Tel: +44 (0) 1223 715000 Fax: +44 (0) 1223 715001

1 0

RE: Low SMB throughput
by Hawley, Rob 28 Sep '00

28 Sep '00

What version of the filer are you running? This is the checklist I give in the cifs performance class I teach for the netapp SE's. This is based on the experience of actual customers in order of decreasing probability. 1. Network Problems. In one of your earlier emails you stated you were sure this did not apply. Since NFS performance is OK, you are probably correct, but I would run netdiag anyway. 2. Virus Scanner. Some of the previous years implementations of virus scanners were not very friendly with file servers. If you are running a virus scanner on the affected clients, then you might want to temporarily disable it to diagnose the problem. 3. Oplock problems. There is a known timing bug in Win95 redirector that the filer triggers. The short version is that the client is not able to immediately receive a response from a file server. The filer has a provision to insert delays in its responses to avoid this problem. Take a look at the cifs stat OpLkBkNoBreakAck to see if this is happening. If you are getting counts on this stat, please contact CS. We can tune the filer for your environment so this will stop happening. 4.Max_mpx too small. Normally I include this in any list. Since this is happening with Win95 clients, this can be eliminated as a source. If you cannot identify the problem based on this list, then please work with NetApp CS. They can perform some more detailed tests to identify the source of your problem. mailto: hawleyr(a)netapp.com US Mail: Network Appliance Inc. Windows Networking Engineering 495 E Java Dr Phone/pager: 408-822-6661 Sunnyvale, CA 94089 http://www.netapp.com -----Original Message----- From: Alfred Lerch [mailto:alfred_lerch@mentorg.com] Sent: Thursday, September 28, 2000 12:07 AM To: Toasters Subject: Low SMB throughput Friends, we are running a F740 in a heterogenous NFS/SMB environment. While we are completely happy on NFS performance SMB performance is not what we would like it to be. But it is only on some Win95 and NT 4.0 where then throughput is low, other Win95 and NT 4.0 systems with the same configuration are fine (as far as you are able to state that two Windows installations are the same). Looking to our 3Com switches shows us that it is not an issue on the ethernet level. I personally guess that it is a protocol issue between the NetApp and Microsoft implementation of CIFS/SMB/NetBIOS over TCP/IP but I'm unable to verify this. Any ideas where we might start looking ? Has anyone experienced a similar issue ? regards alfred -- -- \\\\\/// / ^ _ \ ( (o) (o) ) --- That's all folks --- +=oOOO=====(_)=====OOOo==========================================+ |Alfred Lerch Mentor Graphics Deutschland GmbH| |Network & System Administrator Elsenheimer Str. 41 - 43| |IT - Business Support Team Munich D-80687 Muenchen, Germany| |email: alfred_lerch(a)mentor.com Fax : +49 89 57096-400| |Tel. : +49 89 57096-241 Mobile: +49 172 8915200| | | |Helpdesk IT Central Europe in Munich | |Tel. : +49 89 57096-110 internal: 777| |email: helpdesk_dem(a)mentor.com | +=============Oooo.==============================================+ .oooO ( ) ( ) ) / \ ( (_/ \_)

1 0

Do F740 cards work in an F760
by Matt Phelps 28 Sep '00

28 Sep '00

Folks, I know I can get this answer from NetApp, et. al., but this list is easier ;-) Can we use the cards we have in our 740 (gigabit and fiber channel controllers) for an anticipated 760 upgrade? Thanks, -- Matt Phelps System Administrator, Computation Facility Harvard - Smithsonian Center for Astrophysics mphelps(a)cfa.harvard.edu, http://cfa-www.harvard.edu/~mphelps

2 1

Re: Low SMB throughput
by andrew-m.webster＠db.com 28 Sep '00

28 Sep '00

Hi Chris, Where is the delay? Is it that the inital connection to the Filer takes a long time? This would indicate that the Filer is having problems authenticating users via the domain controller. If not have a look at the output from the following commands:- rc_toggle_basic netdiag It might provide some clues. BTW: We never rely on autonegotiate (AUTO), but set each Filer interface to 100meg/Full Duplex. Regards -- Andrew Webster PC, NetApp Filer, TSM Sanno Move Project Deutsche Real Estate Consulting Ltd. - Tokyo Internal (but outside Tokyo): 88 408-7542 Tel. +81-3-5156-7542 Fax. +81-3--5156-4811 Mobile: 0907 184 8958 Andrew-M.Webster(a)db.com ---------------------------------------- Message History ---------------------------------------- From: alfred_lerch(a)mentorg.com on 28/09/2000 08:23 AM GMT To: chris_good(a)webtop.com; toasters(a)mathworks.com cc: Subject: Re: Low SMB throughput Hi Chris, thanks for your quick answer. - autonegotiation and duplex are fine, no errors reported on the switches - pure TCP/IP is fast (like http, ftp etc.) - the Win95 and NT 4.0 which are slow to NetApps are fast to NT fileservers alfred chris(a)webtop.com wrote: > > In toasters(a)mathworks.com, you wrote: > >Any ideas where we might start looking ? Has anyone experienced a > >similar issue ? > > We currently have similar problems. Check network settings, in > particular autonegotiation and duplex. Also check your WINS servers > and PDC/BDC are functioning ok. > > CHris > -- > Chris Good WebTop.com http://www.webtop.com > Tel: +44 (0) 1223 715000 Fax: +44 (0) 1223 715001 -- -- \\\\\/// / ^ _ \ ( (o) (o) ) --- That's all folks --- +=oOOO=====(_)=====OOOo==========================================+ |Alfred Lerch Mentor Graphics Deutschland GmbH| |Network & System Administrator Elsenheimer Str. 41 - 43| |IT - Business Support Team Munich D-80687 Muenchen, Germany| |email: alfred_lerch(a)mentor.com Fax : +49 89 57096-400| |Tel. : +49 89 57096-241 Mobile: +49 172 8915200| | | |Helpdesk IT Central Europe in Munich | |Tel. : +49 89 57096-110 internal: 777| |email: helpdesk_dem(a)mentor.com | +=============Oooo.==============================================+ .oooO ( ) ( ) ) / \ ( (_/ \_)

2 1

Re: Low SMB throughput
by Alfred Lerch 28 Sep '00

28 Sep '00

Hi Chris, thanks for your quick answer. - autonegotiation and duplex are fine, no errors reported on the switches - pure TCP/IP is fast (like http, ftp etc.) - the Win95 and NT 4.0 which are slow to NetApps are fast to NT fileservers alfred chris(a)webtop.com wrote: > > In toasters(a)mathworks.com, you wrote: > >Any ideas where we might start looking ? Has anyone experienced a > >similar issue ? > > We currently have similar problems. Check network settings, in > particular autonegotiation and duplex. Also check your WINS servers > and PDC/BDC are functioning ok. > > CHris > -- > Chris Good WebTop.com http://www.webtop.com > Tel: +44 (0) 1223 715000 Fax: +44 (0) 1223 715001 -- -- \\\\\/// / ^ _ \ ( (o) (o) ) --- That's all folks --- +=oOOO=====(_)=====OOOo==========================================+ |Alfred Lerch Mentor Graphics Deutschland GmbH| |Network & System Administrator Elsenheimer Str. 41 - 43| |IT - Business Support Team Munich D-80687 Muenchen, Germany| |email: alfred_lerch(a)mentor.com Fax : +49 89 57096-400| |Tel. : +49 89 57096-241 Mobile: +49 172 8915200| | | |Helpdesk IT Central Europe in Munich | |Tel. : +49 89 57096-110 internal: 777| |email: helpdesk_dem(a)mentor.com | +=============Oooo.==============================================+ .oooO ( ) ( ) ) / \ ( (_/ \_)

2 1

(no subject)
by Uros Lampret 28 Sep '00

28 Sep '00

Are there any special considerations on using F740 with VMS? I tried once to mount NFS mount on VMS, there vas no problems, but VMS seems to create some other files. Any experience? _____________________________ Uros Lampret _____________________________

1 0

Re: NFS client performance
by David Lockwood 27 Sep '00

27 Sep '00

We had a very similar situation when we first connected an F760 to HP-UX 10.20. Performance was very poor initially, after some investigation we found that having no (0) biods running we got much better performance. This can be changed on HP-UX (10.20 or 11.00) by changing NUM_NFSIOD in the /etc/rc.config.d/nfsconf file. Nobody could tell me exactly how the system was working with no biod processes but it does. >>> "Rick Hulsey" <RHULSEY(a)wnco.com> 09/27/00 02:01pm >>> We have a couple of HP-UX 11.0 systems running Oracle 8i (8.1.6). Each of these systems has about 10-15 small to medium sized Oracle databases on the system. In order to separate tables, logs, and archives for availability and recovery in the event of a filer volume failure, we are using 3 volumes per database. We are using qtrees allow for the mounting the database tables, logs, and archives separately which allows us to use failover between the HP servers for each database. This all boils down to needing 45+ NFS mounts per client. HP still uses separate biod processes (the default is 4 biod processes) , and I think each running biod can handle 2 concurrent I/O requests (this may or may not be configurable by through some kernel parameter - not sure what the parameter is though). Does anyone know how to tune the HP-UX client for the best performance for this many mounts? I currently think we should bump up the number of biod processes to about 32. We will probably bump up to 16 for a while just as an intermediate test before we go to 32. Does anyone know of other changes that will increase NFS client performance? Rick Hulsey Southwest Airlines 2425 Wyman Mail Stop 4DC Dallas, Tx., 75235 rick.hulsey(a)wnco.com (214) 792-7188 Office (972) 880-6882 Cell (800) 915-3747 Pager ---------------------------------------------------------------------- For further information on Benetton Formula 1 visit our web site at www.benettonf1.com. This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please destroy and notify Benetton Formula on +44 1608678000. ----------------------------------------------------------------------

1 0

Re: NFS client performance
by Herbert Kornfeld 27 Sep '00

27 Sep '00

Well according to Netapp solution id 3.0.2639783.2627617, you should actually _reduce_ the number of biods. There's no reference to a particular version of HP-UX, so I presume it applies to everything. Maybe ;-) --h _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. Share information about yourself, create your own public profile at http://profiles.msn.com.

1 0

NFS client performance
by Rick Hulsey 27 Sep '00

27 Sep '00

We have a couple of HP-UX 11.0 systems running Oracle 8i (8.1.6). Each of these systems has about 10-15 small to medium sized Oracle databases on the system. In order to separate tables, logs, and archives for availability and recovery in the event of a filer volume failure, we are using 3 volumes per database. We are using qtrees allow for the mounting the database tables, logs, and archives separately which allows us to use failover between the HP servers for each database. This all boils down to needing 45+ NFS mounts per client. HP still uses separate biod processes (the default is 4 biod processes) , and I think each running biod can handle 2 concurrent I/O requests (this may or may not be configurable by through some kernel parameter - not sure what the parameter is though). Does anyone know how to tune the HP-UX client for the best performance for this many mounts? I currently think we should bump up the number of biod processes to about 32. We will probably bump up to 16 for a while just as an intermediate test before we go to 32. Does anyone know of other changes that will increase NFS client performance? Rick Hulsey Southwest Airlines 2425 Wyman Mail Stop 4DC Dallas, Tx., 75235 rick.hulsey(a)wnco.com (214) 792-7188 Office (972) 880-6882 Cell (800) 915-3747 Pager

1 0

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 September 2000