toasters

toasters@lists.teaparty.net

2 participants
13529 discussions

Re: Filer Performance in a CIFS environment
by Eyal Traitel 08 Aug '00

08 Aug '00

Just general suggestions: 1. Are you sure this is a CIFS-only problem ? Check that your network cards are set to 100MBit, Full-duplex on both sides, or better yet, move to gigabit... 2. Why not buy an HTTP license and use the filer as the web server itself ? 3. To the point - check your filer OnTAP version - there were many CIFS changes lately as I know - maybe performance also... 4. Be more explicit - give us your "sysstat", "sysconfig -r". You can also run from rc_toggle_basic "statit -b", then after some web pages flowing around, do "statit -e" into a file. You will be able to see disk activity there and see if it's the filer's "behinds". Check also for network errors in "netstat". Eyal. --- Mike Hopkins <Mike(a)broaddaylight.com> wrote: > I know this is a generalized question where the answer is shaped by a > basketful of variables, but I need to start somewhere. I have a > Filer 740 > that I am using to store files for a Win2000 web server. > > Pages served up by this combination of hardware are noticeably > slower(5 to > 10 seconds versus 1 second) than storing the web content on the local > drive > of the web server. This may not seem like much of a slowdown, but > people > notice these things and it is a big issue. > > So, what are some of the things that cause performance hits in a CIFS > environment. Any suggestions, war stories, magic potions, or good > old > fashioned know-how is greatly appreciated. > > > > Mike Hopkins > Broad Daylight, Inc. ===== Yours, Eyal Traitel eTraitel(a)yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! *** __________________________________________________ Do You Yahoo!? Kick off your party with Yahoo! Invites. http://invites.yahoo.com/

1 0

No TIME STAMP for new files created on Filer
by Bawa, Rashid 08 Aug '00

08 Aug '00

> We have a F740 Netapps Filer which has filesystems NFS mounted on to a Sun > Unix server (Solaris 2.7). > > When new files are created on the NFS mounted filesystem on the filer, the > time stamp does not show-up ("ls -la") until later; it looks as though > there is some event that triggers the display of the time-stamp. > > Wondering if anyone has come across the same problem/feature! >

7 9

RE: Moving from one NT domain to another.
by Summers, John 07 Aug '00

07 Aug '00

Jeremy, From a recent training course, it was noted that if you move a filer between NT Domains, the lclgroups.cfg is not refreshed. It is best to delete this file and have it recreated by the filer when joining the new domain. Also be sure to add the filer's host name to the NT Domain (under NT) before running the setup information on the filer. John Summers -----Original Message----- From: Jeremy A. Rosengren [mailto:jeremy@rosengren.org] Sent: Monday, August 07, 2000 1:45 PM To: toasters(a)mathworks.com Subject: Moving from one NT domain to another. We have an interesting problem that I'm hoping somebody on the list has first-hand experience with. We could use some suggestions. We are going to be moving from one NT domain to another. An account in Domain A, user1 does not have the same SID as the same account name in Domain B, so once you move, you would have to go through and reassign access permissions with the account in Domain B. This would be acceptable to us, but the problem is the same situation occurs for Administrator between the two domains, so we're worried about losing Administrator access to our filer. We would be able to put up with reassigning all user permissions afterward if we could resolve the Administrator problem. The question is, how do we move a filer from one NT domain to another, and still have Administrator access to it afterward? Thanks, -- jeremy

1 0

Moving from one NT domain to another.
by Jeremy A. Rosengren 07 Aug '00

07 Aug '00

We have an interesting problem that I'm hoping somebody on the list has first-hand experience with. We could use some suggestions. We are going to be moving from one NT domain to another. An account in Domain A, user1 does not have the same SID as the same account name in Domain B, so once you move, you would have to go through and reassign access permissions with the account in Domain B. This would be acceptable to us, but the problem is the same situation occurs for Administrator between the two domains, so we're worried about losing Administrator access to our filer. We would be able to put up with reassigning all user permissions afterward if we could resolve the Administrator problem. The question is, how do we move a filer from one NT domain to another, and still have Administrator access to it afterward? Thanks, -- jeremy

2 1

W2000 CIFS login problem DOT 5.3.5R2P2
by Steve Losen 07 Aug '00

07 Aug '00

I was wondering if anyone else has seen this and if they know whether it is a problem with W2000 or DOT 5.3.5R2P2 We are using unix authentication for CIFS logins, so W2000 is set up for logging into a samba server. It works fine on a samba server but we are having problems with the filer. We turned on CIFS login logging to /etc/messages and the problem is obvious. The first character of the loginid is missing in the log message. I don't know if the PC or the filer is losing it. We'll have to set up a sniffer to find out. Has anyone else seen this one? I'm on NOW looking for this bug. If I don't find it, I'll open a case. There was an earlier bug where W2000 could not login to CIFS with unix authentication, but that is fixed in 5.3.5R2P2. Perhaps it wasn't completely fixed. We are having no problems with CIFS logins from NT, W95, and W98. (Have to enable clear text passwords, of course.) And I'm pretty sure some folks have had success from W2000, so this probably does not happen to everyone. Steve Losen scl(a)virginia.edu phone: 804-924-0640 University of Virginia ITC Unix Support

1 0

RE: Replacing defective DLT7000 drives without rebooting the fil er
by Kelly Wyatt 07 Aug '00

07 Aug '00

We always power down the filer: Halt filer remove scsi cable from tape adapter boot filer ... 2-3 minute outage replace the drive. Halt filer attach scsi cable from tape adapter boot filer ... 2-3 minute outage Two short outages are easier to handle... yes, we have CIFS and NFS connections that are affected. In theory, since we schedule the outage, the users should be prepared... :-) We currently have 2 9710's but will be consolidating into one L700 -- which has, as you mention, hot swappable drives. But I believe that NetApp still recommends that you power down the filer. Would be nice if that was not so... Kelly -- Kelly Wyatt, Kelly.Wyatt(a)SAS.com Senior Systems Programmer Strategic Recovery SAS Institute Inc. / SAS Campus Drive / Cary, NC 27513 http://www.sas.com SAS... The Power to Know -----Original Message----- From: Micheler Klaus (MDCA Villach) [mailto:Klaus.Micheler@infineon.com] Sent: Monday, August 07, 2000 3:21 AM To: 'toasters(a)mathworks.com' Subject: Replacing defective DLT7000 drives without rebooting the filer We have problems with bad drive heads of DLT7000-drives in our Backup Roboter (Storagetek L700). Regulary about all 2 months one of our Driveheads get defective. Netapp recommends us to power off the filer, replace the drive, and then power on the filer. We don´t like this way, because we have many CIFS-User on that Filer, who will loose their connection in the case of a reboot. My question is how you handle replacing of drives in a library. Do you power the filer down, or do you replace the drive without powering the filer down (if you do so, how???) Our configuration is like follows: F760 Filer with 2 DLT 7000 (Part of an STK L700 Library). Both drives are connected to the filer with one "Fast Wide Differential - SCSI-Interface". The library supports swapping of drives without powering it down. regards, Klaus Micheler Infineon DC Austria

1 0

Re: Virus Scanning
by Thierry.Abbet＠ch.abnamro.com 07 Aug '00

07 Aug '00

Lawrence, We have setup a NT4 server as a ScanServer with McAfee on it. We have a point to point connection between the Filer and the scanserver. (Don't forget to set the speed of the network card on the NT server to 100MB Full Duplex) We just map the cifs volume and scan the cifs volume over the point to point connection(to avoid network load). So far we are very happy with this solution and have been able to detect and remove viruses from the Filer. As soon that a virus is detected a popup window appear on my pc. Have a nice day! ;-) Thierry ABBET IT-Engineering NT / Solaris / VMS System Manager ABN AMRO Bank N.V., Amsterdam Zurich Branch Beethovenstrasse 33 CH-8022 Zurich Switzerland Telephone +41-1-631-5451 ( Direct # ) Telefax +41-1-631-4181 Mobile +41-79-354 3221 Thierry.Abbet(a)ch.abnamro.com (abbet(a)aab.ch) lnail(a)ti.com on 04.08.2000 20:52:31 To: toasters(a)mathworks.com cc: (bcc: Thierry Abbet/CH/ABNAMRO/NL) Subject: Virus Scanning Just curious if anyone is running a virus scanner against filers for CIFS? ______________________________________ Lawrence D. Nail Microsoft Certified Systems Engineer Senior NT Systems Administrator MSPD NT Systems Engineering Mixed Signal Products http://www1.msp.sc.ti.com/nail/ <http://www1.msp.sc.ti.com/nail/>

1 0

Replacing defective DLT7000 drives without rebooting the filer
by Micheler Klaus (MDCA Villach) 07 Aug '00

07 Aug '00

We have problems with bad drive heads of DLT7000-drives in our Backup Roboter (Storagetek L700). Regulary about all 2 months one of our Driveheads get defective. Netapp recommends us to power off the filer, replace the drive, and then power on the filer. We don´t like this way, because we have many CIFS-User on that Filer, who will loose their connection in the case of a reboot. My question is how you handle replacing of drives in a library. Do you power the filer down, or do you replace the drive without powering the filer down (if you do so, how???) Our configuration is like follows: F760 Filer with 2 DLT 7000 (Part of an STK L700 Library). Both drives are connected to the filer with one "Fast Wide Differential - SCSI-Interface". The library supports swapping of drives without powering it down. regards, Klaus Micheler Infineon DC Austria

1 0

RE: Virus Scanning
by Lobban, Robert W SSI-ISEP-3 07 Aug '00

07 Aug '00

Nail, We have two F760 filers running CIFS only. We have about 3000-4000 client connecting to the filers from a number of sites. We are presently using McAfee Virus Scan 4.3.0 to scan both filers. We have set-up several high spec workstation each sweeping a section of the filers. We use snap mirror for disaster Recovery so have a combined disk capacity of about 1.1TB. The Virus scanning works pretty well but we are kind of unhappy at losing the on access scanning functions. Regards Robert Lobban Infrastructure Implementation Shell Services International Ltd, Northern Area Office Loirston House, Wellington Road Aberdeen. AB12 3BH * Office: +44 (0)1224 21 7577 * Fax: +44 (0)1224 21 7501 * Robert.W.Lobban(a)IS.Shell.Com -----Original Message----- From: Nail, Larry [mailto:lnail@ti.com] Sent: 04 August 2000 20:53 To: 'toasters(a)mathworks.com' Subject: Virus Scanning Just curious if anyone is running a virus scanner against filers for CIFS? ______________________________________ Lawrence D. Nail Microsoft Certified Systems Engineer Senior NT Systems Administrator MSPD NT Systems Engineering Mixed Signal Products http://www1.msp.sc.ti.com/nail/ <http://www1.msp.sc.ti.com/nail/>

1 0

RE: Q. on Qtree/Volume/RAID group
by Smart, Steve 05 Aug '00

05 Aug '00

Please read additions to 4 and 5 -----Original Message----- From: Yizhong Zhou [mailto:zhou@mathworks.com] Sent: 04 August 2000 22:16 To: Ng, Pok Cc: 'toasters(a)mathworks.com' Subject: Re: Q. on Qtree/Volume/RAID group On Fri, 4 Aug 2000, Ng, Pok wrote: > I have few questions regarding qtree, volume and raid group > > 1) Is it ture that raid group is added to a volume automatically? A raid group is created automatically when you add new disks to a volume and fill the current raid group. > > 2) is there any command for creating new raid group? > No. When the current raid group is filled, a new raid group is created automatically. > 3) when a qtree is created on a multi raid group volume, does the qtree span > across all raid group? do you have control on which raid group the qtree can > reside on? I don't think you have control over what raid groups a particular qtree can use. > > 4) when the max disk per raid group is increased, does it affect all raid > groups? I believe it only affects the current raid group. AND any subsequent raid groups > > 5) I know if a new disk is added, it will be added to the last raid group of > a volume. Assuming the answer to Q4 is yes, then what happen if the last > raid group of a volume is filled and there exists raid group(not the last > one) in the same volume that is not filled? If the default raid group is increased the current and subsequent groups ONLY can be added to. As raid groups build themselves as they are required, then by default once raid group 3 is started raid group 2 is assumed to be complete. > See 4). Yizhong Zhou zhou(a)mathworks.com, (508) 647-7371 The Mathworks, Inc. 3 Apple Hill Drive, Natick, MA 01760

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