toasters

toasters@lists.teaparty.net

13515 discussions

04 Apr '00

Howdy, At my day job, I've got a F720 running 5.3.4R2. I'm trying to attach a new ADIC FastStor 7 tape library to it. The filer sees the changer and the drive: slot 0: SCSI Host Adapter 0b (Qlogic ISP 1040B) Fast/Wide, Differential Firmware Version 4.65 Clock Rate 60MHz. 1: ADIC FastStor DLT 0115 3: QUANTUM DLT8000 010F But it won't qualify the drive, even if I try using the /etc/cloned_tapes file. disk0> sysconfig -t Tape drive (0b.3) DLT8000 (Non-qualified tape drive) /etc/cloned_tapes: DLT8000 EMULATES DLT7000 (also tried QUANTUM DLT8000 EMULATES QUANTUM DLT7000). Any ideas? Hopefully my registration on now.netapp.com will come thru sometime soon... Thanks, Bryan -- Bryan Fullerton http://bryanfullerton.com/ Core Competency Samurai Consulting Can you feel the Ohmu call?

1 0

Summary: backup recommendations
by Jeff Bryer 04 Apr '00

04 Apr '00

I had asked about software/hardware recommendations to replace our existing Budtool/DLT4000 equipment. This is for backing up NetApp filers (NDMP support a requirement), SUN servers, and a Linux cluster primarily. Ability to scale to include Netware, and NT is also important, as well as the potential to scale to cover campus-wide backups of Win9x/NT/2000, Mac and other clients. Summary of responses: Software: 1) Quick Restore from Workstation Solutions was recommended as the choice most closely resembling Budtool (licensing model, look and feel) and was thought of as having the best NDMP support today. 2) Networker from Legato, good client support but the NDMP support came last (out of the three choices). It was thought that v6 will have much better NDMP support. Plus the NDMP client pack is expected to be a free upgrade to Budtool owners (or maybe restricted to those on maintenance). 3) Netbackup from Veritas. NDMP support rated number 2, good client support. Quite a few people who had used both Legato and Veritas said they had migrated to Veritas (NDMP support being one reason). Hardware: 1) DLT, by far most people recommended using DLT[78]000s. It's established, people trust the reliability. 2) AIT, on paper specs great. Of those who've used it, some have had good results with direct attach to NetApps, others only once the library was configured in DLT emulation mode. The DLT drives may have had more recommendations but that's very much skewed as most people who responded hadn't used AIT drives. I would like to hear from more people who've actually used AIT drivs. Our current thoughts: Software choice: Veritas. The NDMP support today in Legato is not sufficient for our needs. Quick Restore doesn't have support for enough different clients, a smaller user install base, disaster recovery seemed to be more a focus for the upcoming v3 release. Hardware choice: AIT. We've had to replace our DLT7000 drive three times in the last two years, so we don't feel overlly confident about them (our 4000 stackers have been fine). The specs and road map for AIT look good (on paper). This being said, I still would like to hear from AIT users to feel better about the reliability of the media/drives. Thanks to everyone who responded. -- Jeff Bryer bryer(a)sfu.ca Systems Administrator (604) 291-4935 Academic Computing, Simon Fraser University

1 0

Budtool, NetApp filer & Tape library
by Rainchik, Aleksandr (MED, Non GE) 04 Apr '00

04 Apr '00

Hello, We have NetApp F740. For backups we are using Budtool 4.6.1 with two media servers (Sun's) and tape library connected to them. Right now, when we are doing backups, all data bits are flying over the network from the NetApp to the media server and finally to the tape. I'm curious, if we can connect one SCSI channel from the tape library to the NetApp? And a) use it as NDMP media server b) eliminate data transfer over the network while doing NetApp backups. Any hidden traps? So, we are basically want to replace one media server with a NetApp. And my main concerns is how well NetApp is going to work this way. Can I use all those jbmgr commands to load and unload tapes there? Will index tables be updated correctly on main Budtool server? Thank you.

2 1

Re: Messages from cron_admin about daylight saving time
by Chris Thompson 04 Apr '00

04 Apr '00

Back on Tuesday 21 March I wrote: > > We run a default raid scrub regime (01:00 on Sundays). Last Sunday we > got a message > > Sun Mar 19 01:00:00 GMT [cron_admin]: > Rescheduling raid_scrub_admin to run at Sun Mar 19 01:00:00 GMT 2000 > because of Day Light Saving time change > > [Raid scrubbing itself ran at this time with no problem.] > > Moreover, every time timed tweaks the clock (we use ntp synchronisation) > it again mutters. e.g. > > Tue Mar 21 16:04:33 GMT [cron_admin]: > Rescheduling raid_scrub_admin to run at Tue Mar 21 16:04:33 GMT 2000 > because of Day Light Saving time change > > Now all this is no doubt because next Sunday is clock-changing day > in our timezone (GB), with 2000-03-26 01:00:00 GMT -> 02:00:00 BST. > [American friends will no doubt be seeing this effect next week!] Well, they obviously didn't, and I think I understand why. Contrary to my initial impression, it seems that the USA switches at 02:00:00->03:00:00 (local time), rather than 01:00:00->02:00:00. The messages apparently only happen if the (local) time that the raid scrub is scheduled for does not logically exist. This happened in the UK and in the Republic of Ireland (I got a confirmation from Garret Flynn) and presumably in Portugal (any NetApp users there?). Further east in Europe the clock change is later in local time (because it is at the same real time, 01:00:00 UTC). For those interested: the raid scrub *did* run at the expected time on clock-changing night, and the messages then disappeared. Garret tells me that this is a known bug, 13649. Chris Thompson University of Cambridge Computing Service, Email: cet1(a)ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QG, Phone: +44 1223 334715 United Kingdom.

1 0

RE:
by Hawley, Rob 04 Apr '00

04 Apr '00

The filer only supports Microsoft NT domain controllers. There are a number of issues not just that samba does not support UNICODE in the same way NT does. -----Original Message----- From: Igor Schein [mailto:igor@txc.com] Sent: Tuesday, April 04, 2000 10:30 AM To: toasters(a)mathworks.com Subject: Hi, My goal is to to get CIFS on a 740 filer to join to PDC running on Samba. Last time I checked there was a unicode issue which was preventing people from using Samba as PDC for netapp filers. I was wondering if the status has changed and any1 on the list has been successfull in doing so. THanks Igor

1 0

None
by Igor Schein 04 Apr '00

04 Apr '00

Hi, My goal is to to get CIFS on a 740 filer to join to PDC running on Samba. Last time I checked there was a unicode issue which was preventing people from using Samba as PDC for netapp filers. I was wondering if the status has changed and any1 on the list has been successfull in doing so. THanks Igor

1 0

RE: How are you mounting CIFS homedirs with Win9x?
by Uren, Jim 04 Apr '00

04 Apr '00

[snip] > > Thanks for the info. Are there any plans to support variable > expansion a la > SAMBA? For example, sample allows you to define a share using > "%u", which > gets expanded to the username on the SAMBA side. That llows you to put > "\\samba-server\\home" in the logon script, and have SAMBA do the > exapansion for you. > > Tim Yes, this capability is in-plan. The bug report (burt) number for this RFE is 13856.

1 0

Re: EMC Celerra vs NetApp Filer
by Todd C. Merrill 04 Apr '00

04 Apr '00

Kendall Libby wrote: > > right now. Going by units of "1 cpu, 1gb cache, 200gb disk", EMC > > comes in cheaper after a small number of units. For us. YMWV. Bruce Sterling Woodcock wrote: > But these wouldn't be units that would be equal comparison. > What matters is primarily ops/$ and GB/$ for roughly the same Kendall indeed pointed out one "unit" measure, combining performance and storage, that works well *for us*. ops/$ and GB/$ may not be a primary concern for everyone [0], nor may it be a majority concern [1]; I don't think it's fair to make such a blanket statement. There are subjective factors, of course. And, if you ask NetApp for their super-duper maintenance and support plans with bacon and extra cheese, in order to try to narrow the service level gap and compare apples-to-apples with what EMC offers by default, the price gap narrows quite a bit, too. Bruce wrote: > not doubting what you say; I'm just annoyed customers are > comparing the two and yet don't appear to have numbers on > this. [...] > Since we don't know how much all the EMC and related stuff cost, it's > hard to accurately compare the two solutions. I bet it is much higher, [...] > Personally I don't mind who prefers what just so long as all the facts > are on the table. Vendor quotes are often confidential, and info gained from NDA's certainly are, so release of numbers can be a big no-no. In a public forum, we can only quote list prices and independent performance numbers and currently shipping features. But here, as in our personal consumer lives, "nobody pays list price." ;) It's hard to make an orange look like an apple; and, it's equally hard to make a pear look like an apple, too. Vendors in a comparison have to be sqeeuzed into a framework that is not complementary to any of them. Apples-to-apples comparisons are very difficult when the paradigms and company philosophies are vastly different, too. If we were talking about CheapPCDiskDriveVendor #1 and #2, it might come down soley to parameter/$ and we probabaly wouldn't be having this interesting discussion. ;) [0] I ran across this quote this weekend which was rather interesting: "There is hardly anything in the world that some man cannot make a little worse and sell a little cheaper, and the people who consider price only are this man's lawful prey." --John Ruskin [2] [1] I'm with Brian (?) on the ops issue, though. More, more, more! [2] I'm not commenting on either vendor in question specifically. I just thought it was a nice thought to think about since so many people get wrapped up in one thing ($) when making purchases, for business or pleasure. Step back from the tree and look at the forest. Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

2 1

RE: Problem with WINS registration on 5.3.4R3
by Hawley, Rob 04 Apr '00

04 Apr '00

The rule on WINS is that the WINS server must be able to ping every address on the filer. We added the -wins option on ifconfig since WINS will not work properly if the filer has a private net. The option says to exclude this interface from WINS registration. If you rc_toggle_basic you can use the nbtstat command to show the current registration status. If the entries say broadcast instead of WINS, then we were not able to complete registration with the WINS server. As Andrew said, NetApp customer support would be happy to help you solve this problem. -----Original Message----- From: Bond, Andrew [mailto:abond@netapp.com] Sent: Tuesday, April 04, 2000 2:12 AM To: johan.gillquist(a)cn.saab.se Cc: toasters(a)mathworks.com Subject: RE: Problem with WINS registration on 5.3.4R3 Maybe the WINS server is not accessible to the filer (in which case there should be errors in the messages file). Some versions of ONTAP allow you to run "cifs testdc" and this will report whether the name was registered by WINS or by broadcast because the WINS server was not available (I think this shows as "WINS" or "WINS broadcast" respectively). It is also possible to disable wins registration for specific IP address using the -wins flag to ifconfig. Have you opened a call with NetApp support? There are other tools to debug this.... Regards, Andrew > -----Original Message----- > From: owner-dl-toasters(a)netapp.com > [mailto:owner-dl-toasters@netapp.com]On Behalf Of Johan Gillquist > Sent: 04 April 2000 07:16 > To: netapp mailing list > Subject: Problem with WINS registration on 5.3.4R3 > > > Hi all, > > I have have a problem with my F720 running 5.3.4R3, CIFS and NFS. The > toaster does not seem to register itself in WINS. When I run a "cifs > sessions" command it states "WINS servers: xxx.xxx.x....." > and it's the > correct address. When a check the WINS server there is no entry's for > the filer. If I enter a static entry in the WINs database it > works fine > > Suggestions ? > > /Johan >

1 0

RE: Problem with WINS registration on 5.3.4R3
by Bond, Andrew 04 Apr '00

04 Apr '00

Maybe the WINS server is not accessible to the filer (in which case there should be errors in the messages file). Some versions of ONTAP allow you to run "cifs testdc" and this will report whether the name was registered by WINS or by broadcast because the WINS server was not available (I think this shows as "WINS" or "WINS broadcast" respectively). It is also possible to disable wins registration for specific IP address using the -wins flag to ifconfig. Have you opened a call with NetApp support? There are other tools to debug this.... Regards, Andrew > -----Original Message----- > From: owner-dl-toasters(a)netapp.com > [mailto:owner-dl-toasters@netapp.com]On Behalf Of Johan Gillquist > Sent: 04 April 2000 07:16 > To: netapp mailing list > Subject: Problem with WINS registration on 5.3.4R3 > > > Hi all, > > I have have a problem with my F720 running 5.3.4R3, CIFS and NFS. The > toaster does not seem to register itself in WINS. When I run a "cifs > sessions" command it states "WINS servers: xxx.xxx.x....." > and it's the > correct address. When a check the WINS server there is no entry's for > the filer. If I enter a static entry in the WINs database it > works fine > > Suggestions ? > > /Johan >

1 0

Problem with WINS registration on 5.3.4R3
by Johan Gillquist 04 Apr '00

04 Apr '00

Hi all, I have have a problem with my F720 running 5.3.4R3, CIFS and NFS. The toaster does not seem to register itself in WINS. When I run a "cifs sessions" command it states "WINS servers: xxx.xxx.x....." and it's the correct address. When a check the WINS server there is no entry's for the filer. If I enter a static entry in the WINs database it works fine Suggestions ? /Johan

1 0

RE: Legato NetApp Client Pak = weak
by Graydon Dodson 03 Apr '00

03 Apr '00

> Is anyone else using the Legato Client Pak for NetApp? The fact that it > doesn't support the GUI doesn't bother me, but I *do* have a problem that I > have to manually restart the daemon on the filer after every reboot. There > doesn't seem to be a way to start if from the rc file, as you have to type > the filer's password. We have had MAJOR problems using Legato's client pack for NetApp. Both with filler panics and with bad backups. My recommendation would be to avoid it. > Does anyone have insight into whether or not Legato will be expanding this > product? Legato's answer to us was to give us a free license to BudTool. But it's future is far from certain. Thanks. Graydon Dodson grdodson(a)lexmark.com Lexmark International Inc.

1 0

Legato NetApp Client Pak = weak
by Timothy Demarest 03 Apr '00

03 Apr '00

Is anyone else using the Legato Client Pak for NetApp? The fact that it doesn't support the GUI doesn't bother me, but I *do* have a problem that I have to manually restart the daemon on the filer after every reboot. There doesn't seem to be a way to start if from the rc file, as you have to type the filer's password. Does anyone have insight into whether or not Legato will be expanding this product? Ti

1 0

RE: How are you mounting CIFS homedirs with Win9x?
by Mohler,Anissa 03 Apr '00

03 Apr '00

You have hit the issue exactly... basically can't do that in 95 - not on Filers or NT Servers. Here are the two options I explored in my implementation of a filer in a CIFS only environment at my old job. I ended up using option #2 with batch files to ease the pain. I just couldn't get kixtart to run consistently. You can do some searches on MS's web site exact command to add in your batch file to automatically assign the user home directory in the NT Profile. Unfortunately I don't have the batch file myself anymore. Two options: #1 You can use a program like WINSET.EXE and a logon script processor like KiXtart to set the environment variable then map a drive for the 95 user in the same manner he would map the drive for the NT users. (Microsoft's web site has some good KB articles on this). #2 You can manually (or through batch file) set the Home Directory property (in NT User Manager for Domains under User Properties/Profile) to the desired home directory for each user. E.g. You would have to set the Home Directory property for your 95 users to Connect Z: (or whatever letter) to \\filer\username. Then you can use the native NT login script to map the 95 users to NET USE Z: /HOME This topic is pretty well covered on Microsoft's web site. Another option you can try is just place a shortcut to the filer on the users desktop. Then they could doubleclick on the filer and they will see their user home directory at the top level. Not as slick as some of the other options, but it should work as well. Hope this helps. :-----Original Message----- :From: Timothy Demarest [mailto:demarest@arraycomm.com] :Sent: Friday, March 31, 2000 10:05 AM :To: toasters(a)mathworks.com :Subject: How are you mounting CIFS homedirs with Win9x? : : : :I was wondering how everyone is mounting CIFS file systems :from Windows 9x :clients? Having worked through the NIS netgroup issue (Bug :#19922, patched :in 5.3.4R3P2 -- thanks everyone), I'm moving on to getting the :new filer :working with PCs. : :I have a CIFS homedir for each user. The filer shares this directory so :that it appears as the user name. In a Windows NT login :script, I can use :the "%username%" directive to mount the user's home dir as H:. :However, you :can't use "%username%" with Windows 9x clients. : :One option is to set the home directory in the user's :environment profile :(in User Manager), but this requires changing each user's profile. The :logon script is simple and elegant. : :I'd really like to mount the CIFS homedir directly, rather :than having the :user open the drive, then open the folder with their :associated username in :order to access files in their home directory. : :Any hints? : :Tim : :

2 1

Use of Netapp as DMZ "glue" device - possible?
by Jim Klun 7183 03 Apr '00

03 Apr '00

I am contemplating this architecture: INTERNET | CISCO PIX FW ------- DMZ FTP SERVER | | ------------------------------ | | | | NT CLIENT NETAPP The DMZ FTP Server would have an a UDP NFS mount THROUGH THE FIREWALL into the internal network. Inbound guest FTP would be directed there via the Netapp NFS mount. The NT box would have a covering SMB mount over the same filesystem. After a guest client delivers data to the FTP server it would be moved from the in to the out directory where it would be picked up by the internal NT box. Intent: Maintain an Internet-visible FTP server in the DMZ and yet provide easy access for inside servers. Question: 1. Does this make sense? Does anyopne use filers this way or in related ways? 2. Are there known exploits against filers doing UDP NFS as I describe above. Could the Netapp be attacked if the FTP box were hacked? 3. Related question: Can admin access to the filer be to ONLY the console port or ONLY a single interface? Thanks - Jim Klun * James A. Klun Sterling Commerce 4600 Lakehurst Ct. Dublin, Oh 43016 * * jklun(a)stercomm.com 614-793-7183 voice 614-793-7092 fax * * ---- "very like a whale" ---- * * ---- Hamlet, Act 2, Scene 2 ---- *

6 7

EMC Celerra vs NetApp Filer
by Gilles TOURPE 03 Apr '00

03 Apr '00

Hello, Did anybody get in the same site, an EMC Celerra system and a NETAPP F7xx ( at least 740 ) Filer ? I would be interrested in Performances, administration, supports issues ... I'm very curious about the Celerra solution. On the paper it looks smart and really really NAS-oriented ( quite funny for SAN-guys :-) ). Some tests we did we pretty concluant also ( and complex also ... ). Any thinking, remarks ???? Thanx Gilles ------------------------------------------------------------ Gilles TOURPE - System and Network Administrator APS BNP ARBITRAGE - Salle des marches ACTIONS - PARIS - France Phone : (33)-1-40-14-24-03 e-mail : gilles(a)bnpcn.com

2 1

Fwd: NetCache/NetApp Release 3.4
by Jamie Lawrence 03 Apr '00

03 Apr '00

From Bugtraq. -j >Approved-By: aleph1(a)SECURITYFOCUS.COM >Delivered-To: bugtraq(a)lists.securityfocus.com >Delivered-To: BUGTRAQ(a)SECURITYFOCUS.COM >X-Hate: Where do you want to go to die? >Date: Thu, 30 Mar 2000 10:16:08 +0200 >Reply-To: Michal Zalewski <lcamtuf(a)DIONE.IDS.PL> >Sender: Bugtraq List <BUGTRAQ(a)SECURITYFOCUS.COM> >From: Michal Zalewski <lcamtuf(a)DIONE.IDS.PL> >Subject: NetCache/NetApp Release 3.4 >X-To: BUGTRAQ(a)SECURITYFOCUS.COM >To: BUGTRAQ(a)SECURITYFOCUS.COM > >NetCache by Network Appliance (www.netapp.com) is one of the most popular >heavy-duty commercial proxy/cache application. In fact, it's rather poorly >written. We can't see the source code, but, some side effects instead. >There's a lot of them, but I'll try to focus on something called 'internal >requests' - requests that access proxy itself and are handled specially. > >For example, by connecting to proxy server and sending 'GET >http://proxy_server_itself:8080/disk_objects/help', we'll (usually) get >error message with this URL and a few stupid characters appended at the >end of it. It won't happen if you specify anything after 'help' - so I >believe it's something like broken sscanf() used to determine which help >object user wants. But that's not the point. Try appending slash and >approx 10k of 'A' letters to our request. In case of any other request >treated as 'external', it might result only in error message. But in this >case, with something around 9850 characters, our connection to proxy >server is immediately dropped... sounds familiar? I believe it's an >overflow. > >Another way to access it (urm, I mean, cause crash) is something like: >'GET disk_object://xx/AAAAA...'. Btw. I'm wondering there's anything >interesting available to download this way? There are some pictures, >disk_object://xx/help/graphics/help.gif and so on, but I haven't access to >NetCache box to check what else is inside... Aaaah, almost forgotten! Any >file within disk_object hierarchy might be downloaded as-is by appending >'/' to URL - for example, 'disk_object://xx/help/graphics/help.gif/' will >return text/plain dump of this GIF. This means, NetCache fails to classify >this file, so if there's any script or other special object, it won't be >recognized as something 'special'? > >-- > >Just to make everything clear - I haven't shell access to running NetCache >box, so I cannot verify I'm absolutely right - eg. if there's anything >interesting within disk_objects or what extactly happens, but I think >there's absolutely something wrong, and there's no excuse for poor, even >commercial, code. > >#include <stddiscl.h> // Standard Disclaimer applies > >_______________________________________________________ >Michal Zalewski * [lcamtuf(a)ags.pl] <=> [AGS WAN SYSADM] >[dione.ids.pl SYSADM] <-> [http://lcamtuf.na.export.pl] >[+48 22 551 45 93] [+48 603 110 160] bash$ :(){ :|:&};: >=-----=> God is real, unless declared integer. <=-----=

4 5

RE: Budtool v4.6.1 NDMP breaks with DOT 5.3.5P2
by Francett, Robert D (SAIC) 31 Mar '00

31 Mar '00

Our Veritas backups via NDMP also broke on reboot following upgrade to DOT5.3.5P2 from 5.3.4R2 (note Veritas not sanctioning 5.3.5anything yet to my knowledge). Got backups restarted, but not easy nor straight-forward. NetApp looking at it. We're downgrading to 5.3.4R3P2 per NetApp. Bob > ---------- > From: Jeff Krueger[SMTP:jkrueger@qualcomm.com] > Reply To: jkrueger(a)qualcomm.com > Sent: Friday, March 31, 2000 2:09 PM > To: Tim McDonough > Cc: Toaster > Subject: Re: Budtool v4.6.1 NDMP breaks with DOT 5.3.5P2 > > Tim: > > We're in the exact same situation. The first problem is that our tape > library device name is now called "mc0" instead of "spt0". This is > configured in the jbmgr_config file. > > Let us know if you figure this one out. We'll do the same if we figure > something out! =) > > -- Jeff > > > From Tim McDonough on Fri, 31 Mar 2000 13:00:54 PST: > > > >Howdy All: > > > > We were placed on the upgrade treadmill last week and > >upgraded > > our F740 cluster from v5.3.2D1 ( NDMP worked ) -- v5.3.4R3 > > -- v5.3.5P2 ( NDMP no worky ) > > > > Subsequently, our direct attached Budtool NDMP backups > >stopped > > working and timed out with each request. > > > > Anyone else seen this, solved this? > > > > Legato and NetApp are scratching their heads when they're not > > > > pointing their fingers at one another. > > > > Tim > > > > > > >

1 0

RE: Budtool v4.6.1 NDMP breaks with DOT 5.3.5P2
by Melvin, Grant 31 Mar '00

31 Mar '00

Hi Jeff, We've built in backwards compatibility, so that the jbmgr_config file can use "mco" or "spt0" even tough sysconfig -m only reports mc0. Hope this helps. Tim, I'd like to understand your issues with 5.3.5P2. Can you send me the case # logged with either NetApp or Legato support (or both)? Otherwise, please send me the symptoms. Cheers, Grant -----Original Message----- From: Jeff Krueger [mailto:jkrueger@qualcomm.com] Sent: Friday, March 31, 2000 3:10 PM To: Tim McDonough Cc: Toaster Subject: Re: Budtool v4.6.1 NDMP breaks with DOT 5.3.5P2 Tim: We're in the exact same situation. The first problem is that our tape library device name is now called "mc0" instead of "spt0". This is configured in the jbmgr_config file. Let us know if you figure this one out. We'll do the same if we figure something out! =) -- Jeff >From Tim McDonough on Fri, 31 Mar 2000 13:00:54 PST: > >Howdy All: > > We were placed on the upgrade treadmill last week and >upgraded > our F740 cluster from v5.3.2D1 ( NDMP worked ) -- v5.3.4R3 > -- v5.3.5P2 ( NDMP no worky ) > > Subsequently, our direct attached Budtool NDMP backups >stopped > working and timed out with each request. > > Anyone else seen this, solved this? > > Legato and NetApp are scratching their heads when they're not > > pointing their fingers at one another. > > Tim > > >

2 1

Budtool v4.6.1 NDMP breaks with DOT 5.3.5P2
by Tim McDonough 31 Mar '00

31 Mar '00

Howdy All: We were placed on the upgrade treadmill last week and upgraded our F740 cluster from v5.3.2D1 ( NDMP worked ) -- v5.3.4R3 -- v5.3.5P2 ( NDMP no worky ) Subsequently, our direct attached Budtool NDMP backups stopped working and timed out with each request. Anyone else seen this, solved this? Legato and NetApp are scratching their heads when they're not pointing their fingers at one another. Tim

3 2

RE: benchmark
by Paquette, Trevor 31 Mar '00

31 Mar '00

Oh brother.. asked and answered.. "Each vendor will have to be asked what method they use in the real world." The answer is "ask them.." Subject Closed. > -----Original Message----- > From: Jay Orr [SMTP:orrjl@stl.nexen.com] > Sent: Friday, March 31, 2000 7:30 AM > To: Paquette, Trevor > Cc: Karl Swartz; toasters(a)mathworks.com > Subject: RE: benchmark > > Yes, and THAT WAS MY ORIGINAL QUESTION THAT YOU RESPONDED TO - what do > Netapps and Sun do? > > > On Thu, 30 Mar 2000, Paquette, Trevor wrote: > > Each vendor either sticks with powers of 2 (as they should) or they > round > > it. > > Each vendor will have to be asked what method they use in the real > world. > > > > > > > > Yes, we all know the theory, but I'm talking real world. As the > > > thread reads, some vendors use 1000 KB = 1MB, while some don't. I am > > > asking what is implemented. Do the majority of the unix vendors adhere > to > > > the proper base 2 theory of metrics or do they use base 10 combos > > > (i.e. the 1GB = 1000 MB)? > > > > > > On Thu, 30 Mar 2000, Paquette, Trevor wrote: > > > > > > > Here we go.. > > > > > > > > A bit is the smallest unit of data in a computer; has a single > binary > > > value > > > > of either 0 or 1 > > > > A byte is a unit of information that is eight bits long. (half a > byte, > > > or 4 > > > > bits is called a nibble) > > > > > > > > Computer storage is measured in byte multiples; which are based on > > > powers of > > > > 2. > > > > Because humans deal mainly with the decimal system, the resulting > number > > > is > > > > usually 'rounded off' > > > > for humans to better comprehend. > > > > > > > > 1 Kbyte = 1024 bytes = 2^10 bytes = 1,024 > bytes > > > > 1 Megabyte = 1024 Kbytes = 2^20 bytes = 1,048,576 > bytes > > > > 1 Gigabyte = 1024 Mbytes = 2^30 bytes = 1,073,741,824 > bytes > > > > 1 Terabyte = 1024 Gbytes = 2^40 bytes = 1,099,511,627,776 > bytes > > > > 1 Petabyte = 1024 Tbytes = 2^50 bytes = 1,125,899,906,842,624 > bytes > > > > 1 Exabyte = 1024 Pbytes = 2^60 bytes = 1,152,921,504,606,846,976 > bytes > > > > = approx. 1 billion GB > > > > > > Humm... Fun.. Do you know off hand how Solaris and netapp defines > a MB > > > or > > > > > GB? In moving around files I started thinking about this, since > > > depending > > > > > on what tool you use you get file sizes in "bytes" "kbytes" "MB" > or > > > "GB" > > > > > (a la gnu df, solaris df, etc...) > > > > > > > > > > On Wed, 29 Mar 2000, Karl Swartz wrote: > > > > > > > > > > > > Uh, if I can ask a simple (stupid?) question - I was looking > at > > > the > > > > > source > > > > > > > code for postmark and noticed this : > > > > > > > > > > > > > > #define KILOBYTE 1024 > > > > > > > #define MEGABYTE (1000*KILOBYTE) > > > > > > > > > > > > > > Isn't a megabite 1024*1024? > > > > > > > > > > > > It depends on who you ask. Long ago, disk vendors decided that > they > > > > > > could make their disks look bigger if they defined MB to be 2000 > > > > > > sectors (assuming 512 byte sectors, or 8000 sectors for 128 byte > > > > > > sectors), and GB became 1000 MB. Of course they also talked > about > > > > > > unformatted capacity to further inflate the size over reality. > > > > > > > > > > > > In a similar vein, telco people talk about 1K = 1000. The 64K > B- > > > > > > channel in an ISDN line is 64000 bits per second, not 65536. > > > > > > > > > > > > I'm not sure about networking, but I think FDDI and 100base-T > both > > > > > > have a data rate of 100 * 10^6 bits/second, not 100 * 2^20. > (Not to > > > be > > > > > > confused with the signalling rate, which as I recall is 125 * > 10^6 > > > for > > > > > > FDDI with some flavor of 4/5 encoding, and possibly something > > > similar > > > > > > for 100base-T.) > > > > > > > > > > > > What's right for Postmark? Good question. It looks like Jeff > > > picked > > > > > > the standard for disk sizes, but it's not clear that's the > standard > > > > > > most applicable to this case. Aren't standards wonderful? :-) > > > > > > > > > > > > -- > > > > > > Karl Swartz Network Appliance Engineering > > > > > > Work: kls(a)netapp.com http://www.netapp.com/ > > > > > > Home: kls(a)chicago.com http://www.chicago.com/~kls/ > > > > > > > > > > > > > > > > ----------- > > > > > Jay Orr > > > > > Systems Administrator > > > > > Fujitsu Nexion Inc. > > > > > St. Louis, MO > > > > > > > > > > ----------- > > > Jay Orr > > > Systems Administrator > > > Fujitsu Nexion Inc. > > > St. Louis, MO > > > > ----------- > Jay Orr > Systems Administrator > Fujitsu Nexion Inc. > St. Louis, MO

4 3

RE: benchmark
by Paquette, Trevor 31 Mar '00

31 Mar '00

Common decency is taking this to private email.. which I am now going to do. Sorry to bother the rest of this list with crap.. > -----Original Message----- > From: Jay Orr [SMTP:orrjl@stl.nexen.com] > Sent: Friday, March 31, 2000 8:48 AM > To: Paquette, Trevor > Cc: toasters(a)mathworks.com > Subject: RE: benchmark > > > By restating someone's question as an answer, it is not "asked and > answered". I had put to this forum the question of what does Netapps and > Sun do, since there are many Netapp people on this and many people who use > Sun here. > > Your answers have been less then helpful - If you don't know then why did > you reply to the original posting? It was a SIMPLE QUESTION, I didn't ask > for the theory of powers of two, I didn't ask "who should I ask", I didn't > say "how does Trevor feel about this?" I simply posted to this informative > list a simple question of WHAT DOES NETAPP and SUN do in formulating what > they consider MB and GB. > > Please, if you have nothing useful to add to a discussion, don't > interject. That's just common decency... > > > On Fri, 31 Mar 2000, Paquette, Trevor wrote: > > > Oh brother.. > > > > asked and answered.. > > > > "Each vendor will have to be asked what method they use in the real > world." > > > > The answer is "ask them.." > > > > Subject Closed. > > > > > -----Original Message----- > > > From: Jay Orr [SMTP:orrjl@stl.nexen.com] > > > Sent: Friday, March 31, 2000 7:30 AM > > > To: Paquette, Trevor > > > Cc: Karl Swartz; toasters(a)mathworks.com > > > Subject: RE: benchmark > > > > > > Yes, and THAT WAS MY ORIGINAL QUESTION THAT YOU RESPONDED TO - what > do > > > Netapps and Sun do? > > > > > > > > > On Thu, 30 Mar 2000, Paquette, Trevor wrote: > > > > Each vendor either sticks with powers of 2 (as they should) or they > > > round > > > > it. > > > > Each vendor will have to be asked what method they use in the real > > > world. > > > > > > > > > > > > > > Yes, we all know the theory, but I'm talking real world. As the > > > > > thread reads, some vendors use 1000 KB = 1MB, while some don't. I > am > > > > > asking what is implemented. Do the majority of the unix vendors > adhere > > > to > > > > > the proper base 2 theory of metrics or do they use base 10 combos > > > > > (i.e. the 1GB = 1000 MB)? > > > > > > > > > > On Thu, 30 Mar 2000, Paquette, Trevor wrote: > > > > > > > > > > > Here we go.. > > > > > > > > > > > > A bit is the smallest unit of data in a computer; has a single > > > binary > > > > > value > > > > > > of either 0 or 1 > > > > > > A byte is a unit of information that is eight bits long. (half a > > > byte, > > > > > or 4 > > > > > > bits is called a nibble) > > > > > > > > > > > > Computer storage is measured in byte multiples; which are based > on > > > > > powers of > > > > > > 2. > > > > > > Because humans deal mainly with the decimal system, the > resulting > > > number > > > > > is > > > > > > usually 'rounded off' > > > > > > for humans to better comprehend. > > > > > > > > > > > > 1 Kbyte = 1024 bytes = 2^10 bytes = > 1,024 > > > bytes > > > > > > 1 Megabyte = 1024 Kbytes = 2^20 bytes = > 1,048,576 > > > bytes > > > > > > 1 Gigabyte = 1024 Mbytes = 2^30 bytes = > 1,073,741,824 > > > bytes > > > > > > 1 Terabyte = 1024 Gbytes = 2^40 bytes = > 1,099,511,627,776 > > > bytes > > > > > > 1 Petabyte = 1024 Tbytes = 2^50 bytes = > 1,125,899,906,842,624 > > > bytes > > > > > > 1 Exabyte = 1024 Pbytes = 2^60 bytes = > 1,152,921,504,606,846,976 > > > bytes > > > > > > = approx. 1 billion GB > > > > > > > > > > Humm... Fun.. Do you know off hand how Solaris and netapp > defines > > > a MB > > > > > or > > > > > > > GB? In moving around files I started thinking about this, > since > > > > > depending > > > > > > > on what tool you use you get file sizes in "bytes" "kbytes" > "MB" > > > or > > > > > "GB" > > > > > > > (a la gnu df, solaris df, etc...) > > > > > > > > > > > > > > On Wed, 29 Mar 2000, Karl Swartz wrote: > > > > > > > > > > > > > > > > Uh, if I can ask a simple (stupid?) question - I was > looking > > > at > > > > > the > > > > > > > source > > > > > > > > > code for postmark and noticed this : > > > > > > > > > > > > > > > > > > #define KILOBYTE 1024 > > > > > > > > > #define MEGABYTE (1000*KILOBYTE) > > > > > > > > > > > > > > > > > > Isn't a megabite 1024*1024? > > > > > > > > > > > > > > > > It depends on who you ask. Long ago, disk vendors decided > that > > > they > > > > > > > > could make their disks look bigger if they defined MB to be > 2000 > > > > > > > > sectors (assuming 512 byte sectors, or 8000 sectors for 128 > byte > > > > > > > > sectors), and GB became 1000 MB. Of course they also talked > > > about > > > > > > > > unformatted capacity to further inflate the size over > reality. > > > > > > > > > > > > > > > > In a similar vein, telco people talk about 1K = 1000. The > 64K > > > B- > > > > > > > > channel in an ISDN line is 64000 bits per second, not 65536. > > > > > > > > > > > > > > > > I'm not sure about networking, but I think FDDI and > 100base-T > > > both > > > > > > > > have a data rate of 100 * 10^6 bits/second, not 100 * 2^20. > > > (Not to > > > > > be > > > > > > > > confused with the signalling rate, which as I recall is 125 > * > > > 10^6 > > > > > for > > > > > > > > FDDI with some flavor of 4/5 encoding, and possibly > something > > > > > similar > > > > > > > > for 100base-T.) > > > > > > > > > > > > > > > > What's right for Postmark? Good question. It looks like > Jeff > > > > > picked > > > > > > > > the standard for disk sizes, but it's not clear that's the > > > standard > > > > > > > > most applicable to this case. Aren't standards wonderful? > :-) > > > > > > > > > > > > > > > > -- > > > > > > > > Karl Swartz Network Appliance Engineering > > > > > > > > Work: kls(a)netapp.com http://www.netapp.com/ > > > > > > > > Home: kls(a)chicago.com http://www.chicago.com/~kls/ > > > > > > > > > > > > > > > > > > > > > > ----------- > > > > > > > Jay Orr > > > > > > > Systems Administrator > > > > > > > Fujitsu Nexion Inc. > > > > > > > St. Louis, MO > > > > > > > > > > > > > > > > ----------- > > > > > Jay Orr > > > > > Systems Administrator > > > > > Fujitsu Nexion Inc. > > > > > St. Louis, MO > > > > > > > > > > ----------- > > > Jay Orr > > > Systems Administrator > > > Fujitsu Nexion Inc. > > > St. Louis, MO > > > > ----------- > Jay Orr > Systems Administrator > Fujitsu Nexion Inc. > St. Louis, MO

2 1

How are you mounting CIFS homedirs with Win9x?
by Timothy Demarest 31 Mar '00

31 Mar '00

I was wondering how everyone is mounting CIFS file systems from Windows 9x clients? Having worked through the NIS netgroup issue (Bug #19922, patched in 5.3.4R3P2 -- thanks everyone), I'm moving on to getting the new filer working with PCs. I have a CIFS homedir for each user. The filer shares this directory so that it appears as the user name. In a Windows NT login script, I can use the "%username%" directive to mount the user's home dir as H:. However, you can't use "%username%" with Windows 9x clients. One option is to set the home directory in the user's environment profile (in User Manager), but this requires changing each user's profile. The logon script is simple and elegant. I'd really like to mount the CIFS homedir directly, rather than having the user open the drive, then open the folder with their associated username in order to access files in their home directory. Any hints? Tim

1 0

RE: benchmark
by Paquette, Trevor 31 Mar '00

31 Mar '00

Each vendor either sticks with powers of 2 (as they should) or they round it. Each vendor will have to be asked what method they use in the real world. > -----Original Message----- > From: Jay Orr [SMTP:orrjl@stl.nexen.com] > Sent: Thursday, March 30, 2000 1:53 PM > To: Paquette, Trevor > Cc: Karl Swartz; toasters(a)mathworks.com > Subject: RE: benchmark > > > Yes, we all know the theory, but I'm talking real world. As the > thread reads, some vendors use 1000 KB = 1MB, while some don't. I am > asking what is implemented. Do the majority of the unix vendors adhere to > the proper base 2 theory of metrics or do they use base 10 combos > (i.e. the 1GB = 1000 MB)? > > On Thu, 30 Mar 2000, Paquette, Trevor wrote: > > > Here we go.. > > > > A bit is the smallest unit of data in a computer; has a single binary > value > > of either 0 or 1 > > A byte is a unit of information that is eight bits long. (half a byte, > or 4 > > bits is called a nibble) > > > > Computer storage is measured in byte multiples; which are based on > powers of > > 2. > > Because humans deal mainly with the decimal system, the resulting number > is > > usually 'rounded off' > > for humans to better comprehend. > > > > 1 Kbyte = 1024 bytes = 2^10 bytes = 1,024 bytes > > 1 Megabyte = 1024 Kbytes = 2^20 bytes = 1,048,576 bytes > > 1 Gigabyte = 1024 Mbytes = 2^30 bytes = 1,073,741,824 bytes > > 1 Terabyte = 1024 Gbytes = 2^40 bytes = 1,099,511,627,776 bytes > > 1 Petabyte = 1024 Tbytes = 2^50 bytes = 1,125,899,906,842,624 bytes > > 1 Exabyte = 1024 Pbytes = 2^60 bytes = 1,152,921,504,606,846,976 bytes > = > > approx. 1 billion GB > > > > > > > > > -----Original Message----- > > > From: Jay Orr [SMTP:orrjl@stl.nexen.com] > > > Sent: Thursday, March 30, 2000 10:52 AM > > > To: Karl Swartz > > > Cc: toasters(a)mathworks.com > > > Subject: Re: benchmark > > > > > > > > > Humm... Fun.. Do you know off hand how Solaris and netapp defines a MB > or > > > GB? In moving around files I started thinking about this, since > depending > > > on what tool you use you get file sizes in "bytes" "kbytes" "MB" or > "GB" > > > (a la gnu df, solaris df, etc...) > > > > > > On Wed, 29 Mar 2000, Karl Swartz wrote: > > > > > > > > Uh, if I can ask a simple (stupid?) question - I was looking at > the > > > source > > > > > code for postmark and noticed this : > > > > > > > > > > #define KILOBYTE 1024 > > > > > #define MEGABYTE (1000*KILOBYTE) > > > > > > > > > > Isn't a megabite 1024*1024? > > > > > > > > It depends on who you ask. Long ago, disk vendors decided that they > > > > could make their disks look bigger if they defined MB to be 2000 > > > > sectors (assuming 512 byte sectors, or 8000 sectors for 128 byte > > > > sectors), and GB became 1000 MB. Of course they also talked about > > > > unformatted capacity to further inflate the size over reality. > > > > > > > > In a similar vein, telco people talk about 1K = 1000. The 64K B- > > > > channel in an ISDN line is 64000 bits per second, not 65536. > > > > > > > > I'm not sure about networking, but I think FDDI and 100base-T both > > > > have a data rate of 100 * 10^6 bits/second, not 100 * 2^20. (Not to > be > > > > confused with the signalling rate, which as I recall is 125 * 10^6 > for > > > > FDDI with some flavor of 4/5 encoding, and possibly something > similar > > > > for 100base-T.) > > > > > > > > What's right for Postmark? Good question. It looks like Jeff > picked > > > > the standard for disk sizes, but it's not clear that's the standard > > > > most applicable to this case. Aren't standards wonderful? :-) > > > > > > > > -- > > > > Karl Swartz Network Appliance Engineering > > > > Work: kls(a)netapp.com http://www.netapp.com/ > > > > Home: kls(a)chicago.com http://www.chicago.com/~kls/ > > > > > > > > > > ----------- > > > Jay Orr > > > Systems Administrator > > > Fujitsu Nexion Inc. > > > St. Louis, MO > > > > ----------- > Jay Orr > Systems Administrator > Fujitsu Nexion Inc. > St. Louis, MO

3 2

Re: nas storage
by bar-lev benjamin 31 Mar '00

31 Mar '00

We did not really get an answer to the question raised. I am curious too as I am currently evaluating SANergy and I am very impressed with what I see. __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com

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