toasters April 2003

toasters@lists.teaparty.net

95 participants
102 discussions

RE: Security on Win2k
by Bill Bello 04 Apr '03

04 Apr '03

Thanks to all who responded. Pls check this fix in ONTAP 6.4.R1: * 91117 Incorrect permission check when creating a file from CIFS Regards, Bill -----Original Message----- From: Mike Sphar [mailto:mike.sphar@Remedy.COM] Sent: Thursday, March 27, 2003 8:22 PM To: toasters(a)mathworks.com Subject: RE: Security on Win2k This doesn't sound correct to me. Have you actually tested it? In windows network permissions, share permissions are applied first, then individual file or directory ACLs are applied, and the most restrictive permissions always apply. So if a share is read-only to a user, then even if all the files in that share are writeable by everyone, then that user should not be able to update those files or create any new files or directories in that share. I would be very surprised to find that Netapp was not emulating the standard windows file server behavior in this case. -- Mike Sphar - Sr Systems Administrator - Remedy, a BMC Software Company -----Original Message----- From: Bill Bello [mailto:Bill.Bello@hds.com] Sent: Thursday, March 27, 2003 6:46 PM To: toasters(a)mathworks.com Subject: Security on Win2k All, A question has come up in our lab on the subject. If 'read only' is set on the filer for a CIFS share, a user can still have full rights to that share IF security is NOT set in Win2k as well. Has anyone figured a way to make the share secure from the filer only? Thanks in advance. Regards, Bill Bello ATC PreSales Open Systems Support Westchester Office (914) 773-9131 New York Office (212) 827-1314

1 0

Re: DFM paging
by Kumar_M 03 Apr '03

03 Apr '03

I mean - compatible Paging service provider for the pager. Mohan. ----- Original Message ----- From: "Alfred Lim" <alf1650(a)hotmail.com> To: "Kumar_M" <movvamohan(a)hotmail.com>; <toasters(a)mathworks.com>; "George Kahler" <george(a)YorkU.CA> Sent: Thursday, April 03, 2003 8:27 AM Subject: Re: DFM paging > just wondering, wat is the service? > > BK > ----- Original Message ----- > From: "Kumar_M" <movvamohan(a)hotmail.com> > To: <toasters(a)mathworks.com>; "George Kahler" <george(a)YorkU.CA> > Sent: Thursday, April 03, 2003 7:37 AM > Subject: Re: DFM paging > > > > You need a Alpha Numeric Pager and a compatible service. > > We use an alias which includes the pager number and it works pretty well. > > > > Mohan. > > > > ----- Original Message ----- > > From: "George Kahler" <george(a)YorkU.CA> > > To: <toasters(a)mathworks.com> > > Sent: Wednesday, April 02, 2003 2:34 PM > > Subject: DFM paging > > > > > > > Hi, do any folks know how does one go about setting up alarms going to > > > a pager ? > > > > > > I'm working with DFM V2.1 > > > Thanks, > > > George > > > > > > > -------------------------------------------------------------------------- > > ----- > > > George Kahler e-mail: george(a)yorku.ca > > > Sr. Systems Administrator humans: (416) 736-2100 > x.22699 > > > Computing and Network Services machines: (416) 736-5830 > > > Ontario, Canada, M3J-1P3 > > > > > > > > >

1 0

DFM paging
by George Kahler 03 Apr '03

03 Apr '03

Hi, do any folks know how does one go about setting up alarms going to a pager ? I'm working with DFM V2.1 Thanks, George ------------------------------------------------------------------------------- George Kahler e-mail: george(a)yorku.ca Sr. Systems Administrator humans: (416) 736-2100 x.22699 Computing and Network Services machines: (416) 736-5830 Ontario, Canada, M3J-1P3

3 2

quota question
by Charles Bartels 02 Apr '03

02 Apr '03

Hi Folks, I have yet another quota question. Do I need explicit extries in the quotas file for all PC users or can I use a wildcard? The reason I'm asking is that all the PC users get mapped to "pcuser" which is the setting in wafl.default_unix_user. So the result is that *all* the users get counted in the same quota. Thanks. -Charles Bartels

1 0

RE: virus scanner
by Mike Sphar 02 Apr '03

02 Apr '03

When we first set ours up, we had the scan servers deregistering because of network errors due to speed/duplex mismatches. (Seems that no matter how often I tell everyone never use auto-negotiate it still keeps happening...) It's definitely worth double-checking the port settings and switch settings of all the systems involved just in case. In our case when the client deregistered, fixing the network problem didn't get the client to re-register, but rebooting the scan server fixed it. It's not clear to me why the loss of a scan server would increase CPU usage on the filer side. The filer is just sending scan requests to the scan servers. With only one server up it's still sending the same number of requests. Especially with those requests being non-blocking that doesn't make sense to me. In fact, it makes me more suspicious of a network issue. Certain types of network problems can cause significant cpu usage. -- Mike Sphar - Sr Systems Administrator - Remedy, a BMC Software Company -----Original Message----- From: Bui, Marcus [mailto:Marcus.Bui@aiminvestments.com] Sent: Wednesday, April 02, 2003 7:47 AM To: 'Merkin, David S'; 'toasters(a)mathworks.com' Subject: RE: virus scanner David, Yes, the virus scanner is on the same segment using Gig NIC. Also, mandatory scanning is OFF on the Filer. Thanks, Marcus -----Original Message----- From: Merkin, David S [mailto:david.merkin@gs.com] Sent: Wednesday, April 02, 2003 7:39 AM To: 'Bui, Marcus'; 'toasters(a)mathworks.com' Subject: RE: virus scanner Is the virus scanner on the same segment as the filers? Also, what kind of network card are you using in the Virus Scanning machine. David -----Original Message----- From: Bui, Marcus [mailto:Marcus.Bui@aiminvestments.com] Sent: Tuesday, April 01, 2003 10:42 PM To: 'toasters(a)mathworks.com' Subject: virus scanner Hi Toasters, I have two NT servers using TrendMicro ServerProtect as a virus scanner for F760 running ONTAP 6.3.1R1. Sometimes one of the virus scanner server did not registered. I restarted the TrendMicro NT service and it registered with the Filer again. The only change lately was to update the WINS Server on the Filer. Apparently, the Filer was consistently at 100% CPU usage. But after the second virus scanner registered to the Filer, the CPU was reduce to 50%. Any ideas why the virus scanner got 'de-registered' and why would this cause high CPU on the Filer? Error: CIFS: Virus scanner 10.10.10.10 attempted to register but had insufficient privilege. Tue Apr 1 20:45:02 CST last message repeated 5 times thanks, Marcus Bui

1 0

RE: virus scanner
by Bui, Marcus 02 Apr '03

02 Apr '03

David, Yes, the virus scanner is on the same segment using Gig NIC. Also, mandatory scanning is OFF on the Filer. Thanks, Marcus -----Original Message----- From: Merkin, David S [mailto:david.merkin@gs.com] Sent: Wednesday, April 02, 2003 7:39 AM To: 'Bui, Marcus'; 'toasters(a)mathworks.com' Subject: RE: virus scanner Is the virus scanner on the same segment as the filers? Also, what kind of network card are you using in the Virus Scanning machine. David -----Original Message----- From: Bui, Marcus [mailto:Marcus.Bui@aiminvestments.com] Sent: Tuesday, April 01, 2003 10:42 PM To: 'toasters(a)mathworks.com' Subject: virus scanner Hi Toasters, I have two NT servers using TrendMicro ServerProtect as a virus scanner for F760 running ONTAP 6.3.1R1. Sometimes one of the virus scanner server did not registered. I restarted the TrendMicro NT service and it registered with the Filer again. The only change lately was to update the WINS Server on the Filer. Apparently, the Filer was consistently at 100% CPU usage. But after the second virus scanner registered to the Filer, the CPU was reduce to 50%. Any ideas why the virus scanner got 'de-registered' and why would this cause high CPU on the Filer? Error: CIFS: Virus scanner 10.10.10.10 attempted to register but had insufficient privilege. Tue Apr 1 20:45:02 CST last message repeated 5 times thanks, Marcus Bui

1 0

RE: virus scanner
by Merkin, David S 02 Apr '03

02 Apr '03

Is the virus scanner on the same segment as the filers? Also, what kind of network card are you using in the Virus Scanning machine. David -----Original Message----- From: Bui, Marcus [mailto:Marcus.Bui@aiminvestments.com] Sent: Tuesday, April 01, 2003 10:42 PM To: 'toasters(a)mathworks.com' Subject: virus scanner Hi Toasters, I have two NT servers using TrendMicro ServerProtect as a virus scanner for F760 running ONTAP 6.3.1R1. Sometimes one of the virus scanner server did not registered. I restarted the TrendMicro NT service and it registered with the Filer again. The only change lately was to update the WINS Server on the Filer. Apparently, the Filer was consistently at 100% CPU usage. But after the second virus scanner registered to the Filer, the CPU was reduce to 50%. Any ideas why the virus scanner got 'de-registered' and why would this cause high CPU on the Filer? Error: CIFS: Virus scanner 10.10.10.10 attempted to register but had insufficient privilege. Tue Apr 1 20:45:02 CST last message repeated 5 times thanks, Marcus Bui

1 0

virus scanner
by Bui, Marcus 02 Apr '03

02 Apr '03

Hi Toasters, I have two NT servers using TrendMicro ServerProtect as a virus scanner for F760 running ONTAP 6.3.1R1. Sometimes one of the virus scanner server did not registered. I restarted the TrendMicro NT service and it registered with the Filer again. The only change lately was to update the WINS Server on the Filer. Apparently, the Filer was consistently at 100% CPU usage. But after the second virus scanner registered to the Filer, the CPU was reduce to 50%. Any ideas why the virus scanner got 'de-registered' and why would this cause high CPU on the Filer? Error: CIFS: Virus scanner 10.10.10.10 attempted to register but had insufficient privilege. Tue Apr 1 20:45:02 CST last message repeated 5 times thanks, Marcus Bui

2 1

RE: post-migration: sloooow response
by Hill, Aaron 02 Apr '03

02 Apr '03

I would start at the bottom and work my way up. So where is the bottom? Check physical and network layer type stuff. You mention no physical changes, but you have moved to layer 3, so it is a safe bet the problem is in there somewhere. Duplex mismatching maybe? Run netdiag -v on your filer and it will give you a pretty good analysis from the filer perspective. Check switches/routers for dropped packets, overflowing buffers and/or packet shaping. If you want to get down to the lowest layer and are good with packet headers/tcpdump, then do a packet trace on the filer (pktt) to a known good host and then to a known bad host (i.e. your ls hanging machine). I am betting netdiag or you switches/routers turn up something. If nothing shows there, work your way up to NFS/CIFS analysis. Thanks, Aaron -----Original Message----- From: pdunkin(a)lucent.com [mailto:pdunkin@lucent.com] Sent: Wednesday, April 02, 2003 2:39 PM To: toasters(a)mathworks.com Subject: post-migration: sloooow response The next (final??) phase of our network migration is almost complete. The server room and the users' offices have been moved from a Layer 2 network to a Layer 3 network, and everything has a new IP address. There have supposedly been no physical changes -- the same switches and servers and clients are involved as last week. One of the problems we're encountering is intermittent slow response time in interaction with the NetApps. Example: I'm on the PC in my office. It has a session open with a Solaris server that automounts my home directory from a filer. If I cd to my home directory, it may take a countable number of seconds before the prompt comes back. If I try to ls my directory, it may take even longer. Running ls in /home (where *all* the home directories are, hundreds of them) takes so long that I've been giving up before I get any output. The slowness isn't consistent. Some of the Solaris servers haven't had much trouble at all. Some are pretty frustrating. It wasn't this way last week before the migration. Any ideas what may be wrong? Where are the likely places to start troubleshooting? Thanks! -- Patricia Dunkin Lucent Technologies pdunkin(a)lucent.com 600 Mountain Avenue 2F-213 Phone: 908-582-5843 Murray Hill, NJ 07974-0636 Fax: 908-582-3662

1 0

post-migration: sloooow response
by pdunkin＠lucent.com 02 Apr '03

02 Apr '03

The next (final??) phase of our network migration is almost complete. The server room and the users' offices have been moved from a Layer 2 network to a Layer 3 network, and everything has a new IP address. There have supposedly been no physical changes -- the same switches and servers and clients are involved as last week. One of the problems we're encountering is intermittent slow response time in interaction with the NetApps. Example: I'm on the PC in my office. It has a session open with a Solaris server that automounts my home directory from a filer. If I cd to my home directory, it may take a countable number of seconds before the prompt comes back. If I try to ls my directory, it may take even longer. Running ls in /home (where *all* the home directories are, hundreds of them) takes so long that I've been giving up before I get any output. The slowness isn't consistent. Some of the Solaris servers haven't had much trouble at all. Some are pretty frustrating. It wasn't this way last week before the migration. Any ideas what may be wrong? Where are the likely places to start troubleshooting? Thanks! -- Patricia Dunkin Lucent Technologies pdunkin(a)lucent.com 600 Mountain Avenue 2F-213 Phone: 908-582-5843 Murray Hill, NJ 07974-0636 Fax: 908-582-3662

1 0

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 April 2003