toasters

toasters@lists.teaparty.net

13531 discussions

vol0 question
by Jeff Kennedy 05 Feb '02

05 Feb '02

Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really. So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own? Thanks. -- ===================== Jeff Kennedy Unix Administrator AMCC jlkennedy(a)amcc.com

4 6

RE: Slow write performance
by Mike Ball 05 Feb '02

05 Feb '02

Edward, Telnet to the netapp and type "sysstat 1". Let it run for about 20 seconds and send us the output. Also, type "ifconfig -a" and send us the output. Mike -----Original Message----- From: Edward Hibbert [mailto:EH@dataconnection.com] Sent: Tuesday, February 05, 2002 2:24 PM To: toasters Subject: Slow write performance We're seeing some performance problems on an F720 which you guys might be able to help with (even though it's not exactly top of the range nowadays). What we see is: - We're driving it over NFS v3. - We get about 1500 ops/sec out of it, of which one third are writes and two thirds are reads. There aren't many file open/close operations. - The operations are to random locations in large (10GB) files. - The CPU is running about 75%, and the network input and output below the throughput of the link we have to it. We've seen both CPU and network go higher if we do simple copy tests. - Looking it it via pktt trace, something approaching 15% of WRITE operations take long enough for the clients to time out and retransmit (so at least 1 second). None of the READ operations do. - The retransmissions appear to come in bunches. For example we'll see a few seconds where the filer doesn't respond, during which time the retransmissions will come in, then it will wake up and send some responses back. - The rest of the time, the WRITES are very fast (sub-ms). This appears to have worsened recently. We tried a couple of things: - We thought that this might be because the disk had got full and fragmented, so we zapped a bunch of data. - We rebooted. Neither of these seemed to help much. sysstat consistently shows a cache age of 1. This, and the bursty nature of the delays, suggest to me that I'm just hitting it too hard, and there's some kind of periodic cache-flushing operation going on, but do any of you folk have any other suggestions? Edward Hibbert Internet Applications Group Data Connection Ltd Tel: +44 131 662 1212 Fax: +44 131 662 1345 Email: eh(a)dataconnection.com Web: http://www.dataconnection.com

1 0

Slow write performance
by Edward Hibbert 05 Feb '02

05 Feb '02

We're seeing some performance problems on an F720 which you guys might be able to help with (even though it's not exactly top of the range nowadays). What we see is: - We're driving it over NFS v3. - We get about 1500 ops/sec out of it, of which one third are writes and two thirds are reads. There aren't many file open/close operations. - The operations are to random locations in large (10GB) files. - The CPU is running about 75%, and the network input and output below the throughput of the link we have to it. We've seen both CPU and network go higher if we do simple copy tests. - Looking it it via pktt trace, something approaching 15% of WRITE operations take long enough for the clients to time out and retransmit (so at least 1 second). None of the READ operations do. - The retransmissions appear to come in bunches. For example we'll see a few seconds where the filer doesn't respond, during which time the retransmissions will come in, then it will wake up and send some responses back. - The rest of the time, the WRITES are very fast (sub-ms). This appears to have worsened recently. We tried a couple of things: - We thought that this might be because the disk had got full and fragmented, so we zapped a bunch of data. - We rebooted. Neither of these seemed to help much. sysstat consistently shows a cache age of 1. This, and the bursty nature of the delays, suggest to me that I'm just hitting it too hard, and there's some kind of periodic cache-flushing operation going on, but do any of you folk have any other suggestions? Edward Hibbert Internet Applications Group Data Connection Ltd Tel: +44 131 662 1212 Fax: +44 131 662 1345 Email: eh(a)dataconnection.com Web: http://www.dataconnection.com

1 0

6.2, R100, Where?
by Ambrose_Earle＠shamrockfoods.com 05 Feb '02

05 Feb '02

Where can I find information about the features scheduled for DataONTAP 6.2? and the R100? and other new things?

1 0

Using spare disk shelves
by William Liu 05 Feb '02

05 Feb '02

Question: Suppose I have an additional disk shelf that has data on it, can I just easily add it to the daisy chain of disk shelves on the netapp? So when I do a "sysconfig -r" it will show as spares? Keep in mind that these drives aren't empty. I don't see why not since the source file/directory structure comes from the already configured disk shelves. Can someone shed some light on this? Thanks.

2 1

new bug 60490 cifs and symlinks
by neil lehrer 04 Feb '02

04 Feb '02

has anyone seen this issue? i am going to be using 1 symlink per user for homedirs because our homedir struct is not flat - h/a/aboy, h/b/balbert with dirs a - z. the homedir option will point to the homelinks dir which contains all the symlinks. Description Prolonged heavy access by CIFS clients that involve or cause lookup of Unix absolute symbolic links can affect filer performance. Because the resolution of absolute symlinks depends on Unix mount point information which is not available to CIFS clients, a table of translations is kept on the filer (/etc/symlink.translations). The entries in this file are order dependent and are searched sequentially. If a CIFS client is doing symlink lookups frequently, and if the entry to resolve the symlink is near the end of the file, and if the file is large, there can be a great deal of CPU utilization. This can slow operations for other users. For example, a user may be changing files in a directory that contains dozens of absolutes symbolic links. The /etc/symlink.translations file may have 1000s of entries. The user is accessing the files with Windows Explorer. After the target file is updated, the filer notifies Explorer of the file change. This is because Explorer will have posted a "change notify" request with the filer, so that it is alerted when changes occur and can update/redraw its display window. Explorer will touch each file including each of the symbolic links. The access of the symlinks will cause the filer to attempt to resolve them by scanning the /etc/symlink.translations entries. If there are many symlinks, many entries,multiple users with Explorer windows affected by change-notify updates, and if the entries needed to resolve the symlinks are missing or at the end of the /etc/symlink.translations file then filer performance can be affected. Workaround Use the filer's cifs homedir feature (options cifs.home_dir) instead of using a symlink per user when implementing CIFS home directories if there are many users. -- regards +++++++++++++++++++++++++++++++++++++++++++++++++ + Neil Lehrer + + United States International Broadcasting Bureau + System Development Division + + voice 202 619-2524 + fax 202 619-3576 + nlehrer(a)ibb.gov + + " is this crisis an opportunity or just + another grab the fire extinguisher moment?" + +++++++++++++++++++++++++++++++++++++++++++++++++

4 7

RE: adding cifs users homedirs and setting ownership
by jwitham＠takedapharm.com 02 Feb '02

02 Feb '02

We did the migration from an NT fileserver about a year ago. Granted, we didn't go from NFS to CIFS, but maybe our experience will help a little. The way we define the user's homedir is by throwing the UNC in the Home Directory field in User Mangler, then do a 'net use H: /home' in the logon script. When we migrated, we had our webmaster take 5 seconds out of his busy day to write a quick ADSI script that tore through the SAM of the DCs and simply changed \\NTfileserver\username$ to \\filer\username. Then we used robocopy (NT Res kit) to move everything and their permissions to the new filer. If you're going from NFS to NTFS, you'll probably have to script something. For quotas, we use this line in /etc/quotas: * user@/vol/volX/users 3G (Yeah, I know, 3G for every user is huge.) I think this feature requires ONTAP 6.1, so buyer beware. If you have multiple NT domains, you might want to look at caveats of using the CIFS HOMEDIR option. It's nice not to have 1500 separate shares, but may be prohibitive if you have the same username for different users in multiple domains. Good luck! -john John Witham Sr. Data Networking Engineer Takeda Pharmaceuticals North America, Inc. A computer lets you make more mistakes faster than any invention in human history, with the possible exceptions of handguns and tequila. -Mitch Ratcliffe -----Original Message----- From: neil lehrer [mailto:nlehrer@yahoo.com] Sent: Friday, February 01, 2002 11:30 PM To: toasters(a)mathworks.com Subject: adding cifs users homedirs and setting ownership hi, we are going from nfs, unix on solaris to cifs, ntfs on filer. how do folks do the normal creation of new users homedir[we currently map our h: drive from the login script, we don't set the homedir in the users domain account]. from nt gui, from cmdline, using script? and, for quota purposes, how do you set the new users's homedir ownership and acl, in ntfs? na_chown and some other util? i need to set this up for our user support folks and am looking for best ways. thanks ===== neil lehrer ******************************************** * is this crisis an opportunity for growth * or just another fire extinguisher moment? ******************************************** __________________________________________________ Do You Yahoo!? Great stuff seeking new owners in Yahoo! Auctions! http://auctions.yahoo.com This message is for the designated recipient only and may contain privileged or confidential information. If you have received it in error, please notify the sender immediately and delete the original. Any other use of the email by you is prohibited.

1 0

adding cifs users homedirs and setting ownership
by neil lehrer 02 Feb '02

02 Feb '02

hi, we are going from nfs, unix on solaris to cifs, ntfs on filer. how do folks do the normal creation of new users homedir[we currently map our h: drive from the login script, we don't set the homedir in the users domain account]. from nt gui, from cmdline, using script? and, for quota purposes, how do you set the new users's homedir ownership and acl, in ntfs? na_chown and some other util? i need to set this up for our user support folks and am looking for best ways. thanks ===== neil lehrer ******************************************** * is this crisis an opportunity for growth * or just another fire extinguisher moment? ******************************************** __________________________________________________ Do You Yahoo!? Great stuff seeking new owners in Yahoo! Auctions! http://auctions.yahoo.com

1 0

samba smb.conf sample
by neil lehrer 01 Feb '02

01 Feb '02

hi, i loaded samba 2.2 on my solaris wkstation so i could see whether smbcacls might be useful for unix to ntfs on filer migration. but smbcacls won't run without the smb.conf file. the man page would be helpful if i knew what i was doing, but i don't. can someone send me a basic smb.conf that i can modify for my environment? i didn't see any examples at samba.org. thanks. --

1 0

RE: Whos got that file?
by Drew O'Donnell 01 Feb '02

01 Feb '02

Netwatch.exe, It comes with the NT 4.0 Resource kit. Works great. Drew -----Original Message----- From: Knight, Jason [mailto:Jason.Knight@citadelgroup.com] Sent: Friday, February 01, 2002 8:23 AM To: 'Sam Schorr'; Clawson, Simon; toasters(a)mathworks.com Subject: RE: Whos got that file? Two tools that I use are: 1. Winfile.exe (remember Windows 3.1?) 2. Hyena - this tool used by many NT admins works great on filers for managing files/shares. http://www.adkins-resource.com Hyena might save you from writing a perl script or two (if you have 100's of users with 1000's of files open). Jason Knight Citadel Investment Group L.L.C. -----Original Message----- From: Sam Schorr [mailto:sschorr@homestead-inc.com] Sent: Friday, January 25, 2002 9:41 AM To: Clawson, Simon; toasters(a)mathworks.com Subject: RE: Whos got that file? Yes. The command "CIFS sessions" returns the list of clients and the number of files they have open. The command "CIFS sessions clientname" returns the full file path on the filer that clientname has open and the kind of open. --sam -----Original Message----- From: Clawson, Simon [mailto:simon_clawson@mentorg.com] Sent: Friday, January 25, 2002 6:35 AM To: 'toasters(a)mathworks.com' Subject: Whos got that file? Hi! With CIFS Sessions I can see who's got a CIFS seesion open, but is there a way of seeing what file or path they are using? Simon Simon Clawson HDS Team Systems Administrator Mentor Graphics Uk Rivergate London Road Newbury Berkshire RG14 2QB <<Clawson, Simon.vcf>>

1 0

Jump to page:

2026

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