toasters

toasters@lists.teaparty.net

13531 discussions

Using a filer for AFS storage
by frank marxs 09 Apr '03

09 Apr '03

Has anyone succeeded in running AFS while using a Filer as storage? Wondering if anyone has experience running AFS filesystem which use an filer / NFS export as their primary storage ? I want to try AFS exporting a filesystem / volume which is being NFS mounted from a filer ? We have spare filers with lots of spare disk space ... and for a variety of reasons management is insisting that this next project use AFS for more security. I would like to know if I can have my AFS server use the filer for disk space, and give me an AFS export to my AFS clients... Any thoughts on feasibility, or potential problems? In this scenario the Filer would have an export dedicated only to the AFS server [potentially on a private network] and the AFS server would be the only system with any control / access to that NFS mount. NFS export AFS export Filer ------------->SUN / IBM / AFS server -----------> AFS Client Mounting NFS Storage Exporting via AFS Can AFS export a filesystem it's mounting via NFS, or does AFS need some sort of raw or device level control to it's filesystem / partition? Cheers, Frank "Everything should be made as simple as possible, but not simpler." - Einstein _________________________________________________________________ Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

1 0

AFS server from Filer
by frank marxs 09 Apr '03

09 Apr '03

Has anyone here ever succeeded in having an AFS server [OpenAFS or other] use a filer for it's storage? I need to do secure AFS file sharing for a variety of reasons, and am wondering if there's a way I can use the free space I have on my filers as storage for this AFS exports? I'm thinking of NFS exporting this storage to an AFS server and letting the AFS server re-export this data via AFS. Cheers _________________________________________________________________ MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. http://join.msn.com/?page=features/virus

1 0

Gxs EDISwitch using NFS on filer
by Garrett Burke 09 Apr '03

09 Apr '03

Hi, We are invetigating running Gxs (Global Exchange Services) EDISwitch enterprise ver 7.5 on HPUX-11i mounting its DB directory via NFS onto a filer. While we can get read rates of 8M/s and write of 6.5M/s for flat files on a lightly loaded 840c, when trying to load the EDISwitch app it is taking approx 20 mins to load ~2GB of DB info (~1.7MB/s). The majority of the 2GB is split across 7 data files. The files are on a newly created volume of 7x36GB disks with no other data on it. Is anyone running EDISwitch with its DB NFS mounted? If so, what config do you have r.e. mount options on HPUX box, filer config and LAN (Jumbo frames?) Gxs says that this is a supported config, but the above rate makes it unusable. GB ---- Garrett Burke, Service Implementation & Support Manager, Eircomnet. http://www.eircom.net

1 0

assitance
by 98 olo str.tony 09 Apr '03

09 Apr '03

Dear, Naturally, this letter will come to you as a surprise since we have not met, permit me however, I am Mr. tony Francis, a Zimbabwean Farmer. I am contacting you as the need to for me to have a foreign partner in your country has become necessary due to my plans to relocate and establish a private company in that region, as the present political instability in my country Zimbabwe, does not encourage financial investment as the environment is not conducive for investment and most of all security is not insured, just as reported by international media. This problem started when the Zimbabwean government under the leadership of President Robert Mugabe announced a degree on May 31st 2000 to seize more than 849 white-owned farms. So far, more than 1,400 white owned farms have been invaded and confiscated, as well as claiming the properties of the farmers. During these periods, as reported by international media and the CNN, more than 6 white farmers were killed, the government is not stopping at this, as just recently the government have asked all white-farmers to give up their farms to black farmers or risk going to prison. Having lived in this country Zimbabwe for more than 35 years, I have no other country to go to, nor can I trace any of my distant relatives. Therefore, I decided to contact you, for your assistance, so that I can establish a company in your country and relocate there. Also, I must re-iterate the importance of having a citizen of your country to front as the owner of the enterprise, due to the fact that I am a foreigner. For this purpose, I have put in a security deposit all monies from the sale of my farm and farming equipments (including tractors etc) which amounted to US$14.5Million, I will upon receiving your consent prepare every document to enable you receive this money on my behalf, pending my arrival to meet with you in your country so that we can both make further arrangements with regards to establishing the company. As a partner you will be entitled to 20% stake in whichever company that you both decide to establish when I arrive your country. I am waiting urgently for your response. Cheers and God bless. Best Regards Tony Francis Tony FARMS ZIMBABWE _____________________________________________________________ The e-latin email services are free and very powerful. Tell your friends about it.

1 0

assitance
by 98 olo str.tony 09 Apr '03

09 Apr '03

1 0

I want to join this mailing list
by Yusuke Matsuda 09 Apr '03

09 Apr '03

I want to join this mailing list.

1 0

RE: quotas and unix automount
by Conzen, Christoph (C.) 09 Apr '03

09 Apr '03

We have this working in two different modes. In case the users homedir is a qtree in itself then the entry in the quotafile looks like: /vol/user/userxyz tree 5G and if the users homdir is a subdir in a qtree then it looks like: userxyz user@/vol/user 5G -C.Conzen -----Original Message----- From: Charles Bartels [mailto:cbartels@openharbor.com] Sent: 09 April 2003 01:23 To: 'toasters(a)mathworks.com' Subject: quotas and unix automount Hi, Is it possible, with individual user quotas and automount, to have mount points reflect the user's quota? For example: Currently df shows my home directory like this: Filesystem kbytes used avail capacity Mounted on pluto:/vol/vol0/home/cbartels 15728640 9979060 5749570 64% /home/cbartels But my quota is this: pluto>quota report /vol/vol0/home/cbartels K-Bytes Files Type ID Volume Tree Used Limit Used Limit Quota Specifier ----- -------- -------- -------- -------- -------- ------- ------- --------------- user cbartels vol0 home 336836 5242880 3545 - cbartels tree 6 vol0 home 9979304 15728640 390987 - /vol/vol0/home It would be great if /home/cbartels looked like a 5GB (my quota) partition instead of the ~16GB which is the size of the qtree quota. -Charles Bartels

1 0

RE: Legato and NDMP
by Hill, Aaron 09 Apr '03

09 Apr '03

We use legato to backup over NFS mounts and it works perfectly. Note: must be hard mounts, not automounter. In fact, we have just installed and R100 and R150 to do staging via legato. The idea being we suck the data on to the R100 quickly and then move it onto tape at our leisure. Snapvault is going to fit in there for fast retrieval. The bonus is that it can be used just like normal Legato restores using the Disk Backup/NetApp Nearstore module for Legato. Aaron -----Original Message----- From: Matt Musgrove [mailto:matt.musgrove@thales-is.com] Sent: Tuesday, April 08, 2003 7:06 PM To: Gavin Kelman Cc: toasters(a)mathworks.com Subject: Re: Legato and NDMP Gavin Kelman wrote: > We've just received a nice new F825, which we're trying to integrate into > our exsiting Legato Networker backup system. I've been speaking to our > Legato support, but they're not getting very far, and each day that passes > is another day for us without backups. > > We have a Quantum ATL1500 jukebox directly attached (Robot controller and > one drive) to a Sun box, which is a Legato storage node. The 2nd SCSI drive > is connected to the back of our F825, which picks it up fine, and is > configured as an NDMP device in Networker. > > Now, the problem is, Networker complains that it can't establish an NDMP > connection to the F825, even tho NDMP is enabled on the F825, and I've > double-checked the username is root and the password is correct. Attempting > to do an inventory in nwadmin fails on either device in the jukebox. As well as setting the root password under "Client Setup" it needs be be set at "Media > Devices". This caught me out after I changed the root password on my filer last week. > > > Also, NDMP doesn't do incrementals the normal way, so I was thinking of > just backing up the F825 over the NFS mounts, but I hear that's not a > good idea with Legato. I set my schedules to do levels 1,2,3,4. etc in the days after a full (level 0) which works fine for me. You can backup over the NFS mounts, I know some people do that. Matt

1 0

quotas and unix automount
by Charles Bartels 08 Apr '03

08 Apr '03

Hi, Is it possible, with individual user quotas and automount, to have mount points reflect the user's quota? For example: Currently df shows my home directory like this: Filesystem kbytes used avail capacity Mounted on pluto:/vol/vol0/home/cbartels 15728640 9979060 5749570 64% /home/cbartels But my quota is this: pluto>quota report /vol/vol0/home/cbartels K-Bytes Files Type ID Volume Tree Used Limit Used Limit Quota Specifier ----- -------- -------- -------- -------- -------- ------- ------- --------------- user cbartels vol0 home 336836 5242880 3545 - cbartels tree 6 vol0 home 9979304 15728640 390987 - /vol/vol0/home It would be great if /home/cbartels looked like a 5GB (my quota) partition instead of the ~16GB which is the size of the qtree quota. -Charles Bartels

1 0

Peoplesoft cache
by Bui, Marcus 08 Apr '03

08 Apr '03

Hello Toasters, I ran into a weird problem with Peoplesoft (PS). The client run PS from the Citrix Metaframe 1.8 server with the PS database on the Unix Server. The client cache folder resides on F760c (ONTAP 6.3.1R1, Gig NIC). The cache folder has the client's configuration settings when the app is started. With this setup, it takes about 5minutes to load the app and 100MB of traffic exchange between the Filer and the Citrix server. Even though the cache folder only has about 1.5MB of files. I'm not an expert in decoding network traces but it show a lot retransmission packets size of 1.4MB with no data. But when I moved the cache folder to another NT server or Filer F820 (ONTAP 6.3.1R1, Gig NIC), the app loads in 10secs. Any ideas? thanks, Marcus Bui

1 0

Re: converting a directory to a qtree (fwd)
by Haynes, Tom 08 Apr '03

08 Apr '03

I was told this never made it to the list. If it is the case that it made the list, but was blocked at NetApp, please excuse the noise. Most move algorithms are: move(old, new) if (!rename(old, new)) { copy(old, new) if (is_dir(old)) { foreach entry in old { move(old/entry, new/entry) } delete(old) } In pre-6.4 systems, we always failed a rename() for directories and files between qtrees. In 6.4 systems, we allow renames() for files across qtrees, but not directories across qtrees. A rename() results in the inode and none of the data nodes, being modified. Thus a 10G file gets moved as fast or as slow as a 1k file. Allowing the rename() of the dir adds too much complexity in that we would have to handle atomicity, recovery, users modifying subdirs, multiple hard links inside the directory (detecting whether it stayed within the tree or not), and so on. Handling just the files reduces the majority of the IO with a resulting savings in time and space. I.e., the inode is all that changed, so that the snapshot ends up with about 4 dirty blocks and not effectively 2 per data block in the file. BTW: The FSF implementation of mv, as seen in most Linux distros, etc, has an optimization which states that if a directory can not be renamed, then its children can also not be renamed. FSF have removed that in coreutils-4.5.2 such that mv will take advantage of this new feature in OnTap 6.4. --------------------------------------------- Chris Thompson asked: > What happens if the file being rename()'d has multiple hard links? Wouldn't > this result in the inode being linked into more than one qtree? I replied: We don't allow the rename, which forces a copy. This in turn reduces the hard link count. ----- Forwarded message from Steve Losen ----- >From owner-dl-toasters(a)netapp.com Fri Mar 14 06:18:05 2003 X-Authentication-Warning: fred.mathworks.com: majordom set sender to owner-toasters(a)mathworks.com using -f Message-Id: <200303141353.h2EDrhB04815(a)sasha.acc.Virginia.EDU> X-Mailer: exmh version 2.4 06/23/2000 with nmh-1.0.4 To: "Toasters" <toasters(a)mathworks.com> Subject: Re: converting a directory to a qtree In-Reply-To: Your message of "Fri, 14 Mar 2003 09:05:46 GMT." <003101c2ea09$1c676aa0$43c1869f@E689> Date: Fri, 14 Mar 2003 08:53:43 -0500 From: Steve Losen <scl(a)sasha.acc.virginia.edu> Sender: owner-dl-toasters(a)netapp.com Precedence: bulk > Hi, > > Has anyone any experience of converting a directory to a qtree (we want to > make use of the qtree to qtree snapmirror features in DOT 6.2 and greater) > > There's a doc on NOW which gives two ways either do a copy, which requires > enough free space (which we don't have) or to use the built in ndmpcopy > feature of DOT. > > There's also the possibility of doing a 'mv' on each file, although I'd be a > bit concerned that this might leave the filesystem in a less than optiomal > state on the filer. > > I'd be interested in hearing from anyone who has done this, especially about > throughput (The directories we need to convert have approx 150G of mail in > them, so lots of small files, on a 840C) > GB > > ---- > Garrett Burke, > Service Implementation & Support Manager, > Eircomnet. > http://www.eircom.net As far as I know, there is no way to move a file across a qtree boundary other than to copy the file and delete the original. Even if you use the unix "mv" command, when crossing a qtree boundary mv must copy and delete. Since you have limited free space, you may need to copy and delete each mailbox individually. If you are using snapshots, this will obviously gobble up most if not all of your snapshot reserve, so you may need to delete all your snapshots beforehand. Steve Losen scl(a)virginia.edu phone: 434-924-0640 University of Virginia ITC Unix Support ----- End of forwarded message from Steve Losen ----- -- Tom Haynes, ex-cfb thomas(a)netapp.com

1 0

Off topic: announcing NetVault list
by Alex Harkema 08 Apr '03

08 Apr '03

> Hi all, > If you are using Bakbone's backupsoftware 'NetVault' you might be interested in this new mailinglist we started today. It's intended to be a knowledge resource and sharing possibility for NetVault administrators. > The list (and all commands) is available via majordomo(a)knowhow.vertis.nl > (list name is list-nvusers). > To subscribe send mail to majordomo(a)knowhow.vertis.nl with in the message body 'subscribe list-nvusers' You can send and will receive messages to/from list-nvusers(a)knowhow.vertis.nl > Kind regards, > Alex Harkema

1 0

Legato and NDMP
by Gavin Kelman 08 Apr '03

08 Apr '03

We've just received a nice new F825, which we're trying to integrate into our exsiting Legato Networker backup system. I've been speaking to our Legato support, but they're not getting very far, and each day that passes is another day for us without backups. We have a Quantum ATL1500 jukebox directly attached (Robot controller and one drive) to a Sun box, which is a Legato storage node. The 2nd SCSI drive is connected to the back of our F825, which picks it up fine, and is configured as an NDMP device in Networker. Now, the problem is, Networker complains that it can't establish an NDMP connection to the F825, even tho NDMP is enabled on the F825, and I've double-checked the username is root and the password is correct. Attempting to do an inventory in nwadmin fails on either device in the jukebox. Also, NDMP doesn't do incrementals the normal way, so I was thinking of just backing up the F825 over the NFS mounts, but I hear that's not a good idea with Legato. We're running DOT 6.3.1r1 and Legato Networker 6.1.3. Anyone have any suggestions, apart from "stop using Legato"? Cheers, Gavin. -- Gavin Kelman Phone : 07967 360 822 Unix Administrator Email : gavin(a)metahusky.net

2 1

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

quota question
by Charles Bartels 01 Apr '03

01 Apr '03

Hi, I have a question regarding quotas. I am accessing my home directory from both the UNIX and PC side of things and I have a number of entries in my quota file including a default quota. # Operations folks cbartels user@/vol/vol0/ops_home 7GB - * user@/vol/vol0/ops_home 3GB - When I do a "quota report" I see 2 lines which seem to refer to my account. The second line appeared after I edited and saved a file on the PC side. It seems like there are 2 separate quotas, one for unix files and one for pc files. I would rather have a single quota for all the files in the user's directory. Is there a way to do that? jupiter> quota report K-Bytes Files Type ID Volume Tree Used Limit Used Limit Quota Specifier ----- -------- -------- -------- -------- -------- ------- ------- --------------- user cbartels vol0 ops_home 1211688 7340032 10444 - cbartels user \cbartel vol0 ops_home 52 3145728 1 - Thanks. -Charles Bartels

2 2

converting a directory to a qtree
by Garrett Burke 01 Apr '03

01 Apr '03

Hi, Has anyone any experience of converting a directory to a qtree (we want to make use of the qtree to qtree snapmirror features in DOT 6.2 and greater) There's a doc on NOW which gives two ways either do a copy, which requires enough free space (which we don't have) or to use the built in ndmpcopy feature of DOT. There's also the possibility of doing a 'mv' on each file, although I'd be a bit concerned that this might leave the filesystem in a less than optiomal state on the filer. I'd be interested in hearing from anyone who has done this, especially about throughput (The directories we need to convert have approx 150G of mail in them, so lots of small files, on a 840C) Thanks, GB ---- Garrett Burke, Service Implementation & Support Manager, Eircomnet. http://www.eircom.net

6 6

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