toasters

toasters@lists.teaparty.net

13515 discussions

RE: File name length and veritas
by Clawson, Simon 01 May '02

01 May '02

I have the last inum that was being written before the failure, so I attempting to run a find in the directory it was working on. However that fails too. The veritas guys were not clear on how this command should be run. I am currently using a mounted directory on a Unix client. Will this be able to find a specific inode on the Netapps? Simon -----Original Message----- From: Steve Losen [mailto:scl@sasha.acc.virginia.edu] Sent: 01 May 2002 14:55 To: Clawson, Simon Cc: 'toasters(a)mathworks.com' Subject: Re: File name length and veritas > Hi! > > I have been told by veritas that I have a file name or path that is longer > than 1024 characters. Apparently this will cause problems with NDMP > > They have told me to do the following :- > find . -inum 5942788 -print > > But this just gives an error due to the value being to great. Anyone seen > this before? I know this might seem like UNix 101..so sorry.. Unix typically limits a path name (filename/filename/filename/ ...) to 1024 bytes. However, it is possible and entirely legal to create directory structures that are so deep that the absolute path name to a file is longer than 1024. To get around this, you usually "cd" down into the directory tree and use relative path names. But many dump utilities insist on listing each file as an absolute path name, and this is usually limited to 1024 bytes. Evidently "find" is also unhappy about your long file name. If you have any deep directory trees, or any directories with long names, then I suggest that you cd into the directory and run find . -inum ... from there. It's possible that someone may have inadvertently created a very deep directory tree due to a programming error of some sort. It very easy to do: while :; do mkdir trouble cd trouble done This will create trouble/trouble/trouble/trouble/ ... until the user hits a disk quota or the file system runs out of resources. But before that happens, this could go thousands of directories deep. Steve Losen scl(a)virginia.edu phone: 434-924-0640 University of Virginia ITC Unix Support

2 1

https vs. http
by Piotr KUCHARSKI 01 May '02

01 May '02

Is there any way to disable http, so only https works? It's easy to enable ssh and disable telnet, but how to accomplish similar for na_admin with http (I have no http licenses, only na_admin). p.

1 0

RE: File name length and veritas
by Clawson, Simon 01 May '02

01 May '02

Hi Levent, Sorry, can you explain what you mean by nlist entries? Thanks Simon -----Original Message----- From: Levent KAPTANOGLU [mailto:lkaptan@prolink.com.tr] Sent: 01 May 2002 14:53 To: Clawson, Simon; toasters(a)mathworks.com Subject: RE: File name length and veritas I dont know about the file names but as far as I know there is a limitation for nlist entries. Maximum of 10240 nlist entries supported -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Clawson, Simon Sent: Wednesday, May 01, 2002 4:00 PM To: 'toasters(a)mathworks.com' Subject: File name length and veritas Hi! I have been told by veritas that I have a file name or path that is longer than 1024 characters. Apparently this will cause problems with NDMP They have told me to do the following :- find . -inum 5942788 -print But this just gives an error due to the value being to great. Anyone seen this before? I know this might seem like UNix 101..so sorry.. Thanks Simon Simon Clawson TEL:- +44 (0)1635 811409 HDL Designer Series Team Systems Administrator FAX:- +44 (0)1635 810108 Mentor Graphics Ireland Ltd (UK Branch) MOB:- +44 (0)7788 716071 Rivergate London Road Newbury Berkshire RG14 2QB

1 0

File name length and veritas
by Clawson, Simon 01 May '02

01 May '02

Hi! I have been told by veritas that I have a file name or path that is longer than 1024 characters. Apparently this will cause problems with NDMP They have told me to do the following :- find . -inum 5942788 -print But this just gives an error due to the value being to great. Anyone seen this before? I know this might seem like UNix 101..so sorry.. Thanks Simon Simon Clawson TEL:- +44 (0)1635 811409 HDL Designer Series Team Systems Administrator FAX:- +44 (0)1635 810108 Mentor Graphics Ireland Ltd (UK Branch) MOB:- +44 (0)7788 716071 Rivergate London Road Newbury Berkshire RG14 2QB

2 1

Multistore/vfiler experiences?
by Brian Tao 30 Apr '02

30 Apr '02

Does anyone have any real-world experience with vfilers (the "MultiStore" capability in 6.2) they can share? I have an opportunity to consolidate a bunch of older F740s onto a pair of clustered F840s, but I'm not sure about putting all my eggs (well, half my eggs) into one basket. It isn't quite possible to get 100% availability with Netapps yet, and it's been nice in the past having a handful of discrete clustered F740 systems for limiting the effect of downtime. The vfiler stuff looks intriguing, though... -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

1 0

Amazon gift certificate
by Maglinger, Paul 30 Apr '02

30 Apr '02

I did a survey for NetApps and got an email from Amazon saying I had a gift certificate but it was blank. Anybody else run into this? Remember, you are unique... just like everyone else! Paul Maglinger, A+, CA, CET, MCSE Systems Administrator Shoe Carnival Inc. (812)867-4674 pmaglinger(a)scvl.com

1 0

Filer Storage Charge Back?
by Infantino, Vincent (Exchange) 30 Apr '02

30 Apr '02

Are there any good utilities (other than quota reports) that NetApp provides or any recommended 3rd Party application that can be used to create a monthly charge back report based on storage allocated to a particular user? Vinny *********************************************************************** Bear Stearns is not responsible for any recommendation, solicitation, offer or agreement or any information about any transaction, customer account or account activity contained in this communication. ***********************************************************************

1 0

Filer takeover
by Jose Celestino 30 Apr '02

30 Apr '02

Hello, I fail to understand the following sentence: "When you configure virtual interfaces for takeover, you must specify the interface name and not the IP address." I have one vif in each filer, separate ones, trunk1=192.168.1.103 on 1st filer and trunk1=192.168.1.104 on 2nd filer. Does the previous sentence mean I have to put: ifconfig trunk1 192.168.1.103 mediatype 100tx-fd partner trunk1 ? That doesn't make much sense. Tried also ifconfig trunk1 192.168.1.103 mediatype 100tx-fd partner 192.168.1.104 but keep getting partner 192.168.1.104 (not in use). TIA. -- Jose Celestino <japc(a)co.sapo.pt> Systems::SAPO.pt http://www.sapo.pt --------------------------------------------------------------------- Titanic 1912 / Hindenburg Zeppelin 1937 / Microsoft Windows 2000 OpenBSD claudia 3.0 GENERIC#94 i386

1 1

RE: Tape drive recommendations
by Peter Bryant 26 Apr '02

26 Apr '02

*** Before acting on this e-mail or opening any attachment you are advised to read the disclaimer at the end of this e-mail *** Whilst it's not here yet, decision making should allow for O-MASS technology that should be out (first generation) in 2003. This is a step change that I've been waiting about 18 months for * tape is much shorter therefore read times much faster * silicon head gives much closer tracks (and is less susceptible to alignment errors due to heat expansion) * tape is native 600GB capacity * 64MB per second transfer * third generation product is predicted at 2.4TB per tape, and 256MB per second <g> Obviously still to get here, but if (like me) you have a large DLT platform and were thinking of using SDLT as a compatible upgrade, it might be worth hanging on with DLT8000 drives, and change to O-MASS in a year or two. Have a look here: http://www.tandberg.com/o-mass/ Peter -----Original Message----- From: Ben Rockwood [mailto:BRockwood@homestead-inc.com] Sent: 25 April 2002 22:53 To: Jeff Kennedy; dwald(a)kersey.yi.org Cc: toasters(a)mathworks.com Subject: RE: Tape drive recommendations I can't resist to step in here. The battle of the tape format is continuing day by day. Many vendors are pushing people away from current DLT technology saying it's doomed, and that AIT isn't far behind. The big pushes are on LTO's and newer drives. I've had my eye on StorageTek for awhile now because they are doing some really neat things, namely FC to the drive in their T9840B drive. The T9940's are nice two with 60G native and 120G 2:1 compressed. But then it all comes down to your stratagy. If you have a small enviroment and you only want to back up, say 1-2TB, then you might like some of the 20slot AIT2/3 arrays from ADIC. SDLT is looking good with native capacities for 110G and higher with compression plus descent speed, however a key factor between DLT and AIT right now is also physical size. If you have a cramped facility you may not have adiquate room for DLT robots, which is where the nice and tiny AITs are handy. There are alot of choices. Currently I'm still liking LTO because it has 15MB/sec sustained rates per drive plus a 100G native capacity, but right now it probly comes down to cost more than anything. In which case, if you buy a robot, buy bigger than you need to so you can scale later. As for which is best specifically for a Filer, I know that you supposedly can put a robot directly onto a head, but I haven't seen anyone do it, and probly would opt to use a seperate gig-network with a dedicated backup server and anything I wanted on the back end of it. Maybe that helps. I hope. Ben Rockwood -----Original Message----- From: Jeff Kennedy [mailto:jlkennedy@amcc.com] Sent: Thu 4/25/2002 6:26 AM To: dwald(a)kersey.yi.org Cc: toasters(a)mathworks.com Subject: Re: Tape drive recommendations Here are my thoughts: AIT: pros: Great product, nice access time due to the embedded chip, best choice for nearline tape storage, small footprint cons: only media supplier is Sony, the rest are just re-sellers (chances are they aren't going anywhere but you never know) SDLT: pros: Long track record, can read older DLT media (if BRC drive), supported by everyone cons: potential for mixing media LTO: not a big fan of LTO, thought it would be great in the beginning but the "unified format" fractured long before I thought it would. ~JK dwald(a)kersey.yi.org wrote: > > Hi all. > > We've been using DLT7000 drives off the back of our 740's for some > time now, and are looking to upgrade. > > According to the NOW site, NetApps supports the SuperDL220, AIT-3 and > some of the LTO's, among others. > > Capacity and throughput are somewhat close across the three types, I'm > curious for any thoughts on which is preferred? Any horror stories on > any of these with Filers? > > Thanks -- ===================== Jeff Kennedy Unix Administrator AMCC jlkennedy(a)amcc.com ******************************* Disclaimer ***************************** Confidentiality: This e-mail and any attachments are intended for the addressee(s) only and may be confidential. If they have come to you in error you must take no action based on them, nor must you copy or show them to anyone; please reply to this e-mail and highlight the error. Viruses: Although we have taken steps to ensure that this e-mail and attachments are free from any virus, we advise that in keeping with good practice the recipient should ensure they are actually virus free. ********************** http://www.burallplastec.com ********************

1 0

Re: Request for Input on Independant Filer tools
by Alek O. Komarnitsky (N-CSC) 26 Apr '02

26 Apr '02

> From owner-toasters(a)mathworks.com Thu Apr 25 18:03 MDT 2002 > > Hello Mighty Toasters, > > As some may have noticed recently, I've become very interested in > monitoring tool development for the filers. I've recently had > a minor breakthrough using SNMP tables and mutli-dimensional arrays > in PERL. This is particularly interesting because it gives > me the ability to develope tools that can run sorts on the raid tables, > disk usage tables, and more. > > Here's the problem, most of the tools I'm building now have little to no value. > They are amusing and fun to write, but especially if you use FilerView these tools are useless. While this is a credit to NetApp, I'm curious if there are any tools people would like to see. > > What types of monitoring tools would you guys like to see? Performance > tools are pretty well taken care of with MRTG and Cricket, though it would > be fairly simple to use an snmp polling loop to drop files into either a > MySQL/PostGreSQL backend or maybe just a BerkleyDB and then run GNUplot on > 'em for a more simplified Cricket replacement, but since Cricket is so trusted > it's unlikely for any of my tools be replace it. > > In essence, I've solved the problems I need to solve and now what to find > some way to use all the kool tricks I've learned in the proccess. Ideas? > > If nothing else, I'm hoping to write an article soon about how to easily > develop your own in-house tools for the Filers, which is probly the most > useful thing I could do. > > Thanx in advance. > > Ben Rockwood It's not very high-tech, but I think syslog analysis from any machines (including filers) can by QUITE helpful. Simply centrally logging events, using a log-parsing tool such as Swatch, and then appplying appropriate filtering rules over time results in a pretty decent signal-noise ratio of "significant" events that merit watching. You still need some other stuff (ex: simple remote monitoring via ping/CIFS/NFS functionality), and doesn't help you too much in the performance area, but syslog is good stuff IMHO. alek P.S. Here's a presentation on using Swatch: http://www.komar.org/ -> Misc. Tech Stuff -> Swatch

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