toasters October 2003

toasters@lists.teaparty.net

80 participants
83 discussions

RE: ONTAP simulator http license?
by Meyer, Brad 28 Jan '04

28 Jan '04

Try these... Brad BKHEXNB is a fcp Simulator site license NAZOMKC is a http Simulator site license -----Original Message----- From: CTG - Bob Weiler [mailto:bweiler@ctgwi.com] Sent: Thursday, June 26, 2003 7:24 AM To: toasters(a)mathworks.com Subject: ONTAP simulator http license? Does anyone know the simulator license for http? When I installed the simulator I noticed the keys for fcp and http are identical. The installation guide has the following: ... license dafs=TBPGWWF license fcp=BKHEXNB <- license http=BKHEXNB <- license nfs=BQOEAZL ... Thanks, Bob -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Dave Hitz Sent: Wednesday, June 18, 2003 8:29 PM To: toasters(a)mathworks.com Subject: ONTAP Simulator available on NOW Some of you may know that NetApp Engineering uses an ONTAP Simulator as part of our development process. (The simulator runs all of ONTAP as a process under Linux. At a very low level, we spoof the disk drivers, so that disks are simulated by local files. We spoof the network drivers by putting the Linux Ethernet card into promiscuous mode and creating a fake MAC address associated with the simulator.) Enough technical details. The point of this message is that we are now making the simulator available to customers so that they can experiment with ONTAP features and commands without putting production systems at risk. I have had many customers tell me that they wished they had a way to play with new releases and new features, but they didn't have a safe place to do it. Just to be clear: This is not production software. It's ONTAP running on top of Linux, so don't use it for performance testing. We have no formal support program. Don't call NetApp support. This is an "as is" tool. (Some restrictions may apply. :-) On the other hand, I think you will find the simulator to be very useful for experimenting with ONTAP. Hundreds of NetApp engineers use it every day, and we have been using the simulator for over a year in NetApp training classes. Not only can you experiment with the features of a single filer, but you can run multiple simulators and experiment with features like SnapMirror, SnapVault, and Cluster Fail Over. It doesn't support Fibre Channel, because that would take more hardware, but it does support iSCSI and all of the other TCP/IP protocols, including NFS, CIFS, HTTP and FTP, as well as network management protocols like SNMP and NDMP. Here are some details: You can download the simulator from NOW at <http://now.netapp.com/NOW/cgi-bin/simulator>. We have created a category in the NOW community forums called "ONTAP Simulator" that can be accessed at <http://forums.netapp.com/showcategory.asp?cid=172>. And finally, we have created a majordomo mailing list that people can join at xdl-simulator(a)netapp.com. (You know the drill. To join send to majordomo(a)netapp.com, not to the simulator list itself.) And of course, feel free to talk about it here on toasters. As I said above, the simulator is NOT formally supported, but some volunteers at NetApp have agreed to informally answer questions to the xdl-simulator mailing list. Have fun! Dave Hitz Founder and EVP Engineering, Network Appliance

2 1

SNMP values keeps changing after every reboot. any solution to fix this?
by Sto Rage© 28 Jan '04

28 Jan '04

Every time the filer is rebooted. the snmp values for disk changes. This affects our mrtg monitoring. Is there a script that helps overcome this issue? How are others monitoring disk usage using mrtg? TIA After last reboot: NETWORK-APPLIANCE-MIB::dfFileSys.1 = STRING: "/vol/vol1/" NETWORK-APPLIANCE-MIB::dfFileSys.2 = STRING: "/vol/vol1/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.3 = STRING: "/vol/vol7/" NETWORK-APPLIANCE-MIB::dfFileSys.4 = STRING: "/vol/vol7/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.5 = STRING: "/vol/vol6/" NETWORK-APPLIANCE-MIB::dfFileSys.6 = STRING: "/vol/vol6/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.7 = STRING: "/vol/vol2/" NETWORK-APPLIANCE-MIB::dfFileSys.8 = STRING: "/vol/vol2/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.9 = STRING: "/vol/vol0/" NETWORK-APPLIANCE-MIB::dfFileSys.10 = STRING: "/vol/vol0/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.11 = STRING: "/vol/vol5/" NETWORK-APPLIANCE-MIB::dfFileSys.12 = STRING: "/vol/vol5/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.13 = STRING: "/vol/vol4/" NETWORK-APPLIANCE-MIB::dfFileSys.14 = STRING: "/vol/vol4/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.15 = STRING: "/vol/vol3/" NETWORK-APPLIANCE-MIB::dfFileSys.16 = STRING: "/vol/vol3/.snapshot" Before the reboot: #NETWORK-APPLIANCE-MIB::dfFileSys.1 = STRING: "/vol/vol2/" #NETWORK-APPLIANCE-MIB::dfFileSys.2 = STRING: "/vol/vol2/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.3 = STRING: "/vol/vol1/" #NETWORK-APPLIANCE-MIB::dfFileSys.4 = STRING: "/vol/vol1/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.5 = STRING: "/vol/vol4/" #NETWORK-APPLIANCE-MIB::dfFileSys.6 = STRING: "/vol/vol4/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.7 = STRING: "/vol/vol6/" #NETWORK-APPLIANCE-MIB::dfFileSys.8 = STRING: "/vol/vol6/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.9 = STRING: "/vol/vol0/" #NETWORK-APPLIANCE-MIB::dfFileSys.10 = STRING: "/vol/vol0/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.11 = STRING: "/vol/vol5/" #NETWORK-APPLIANCE-MIB::dfFileSys.12 = STRING: "/vol/vol5/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.13 = STRING: "/vol/vol3/" #NETWORK-APPLIANCE-MIB::dfFileSys.14 = STRING: "/vol/vol3/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.15 = STRING: "/vol/vol7/" #NETWORK-APPLIANCE-MIB::dfFileSys.16 = STRING: "/vol/vol7/.snapshot"

3 2

RE: Data ONTAP Upgrade
by Sam Schorr 04 Nov '03

04 Nov '03

For all the inquiring minds, I found this under the download section: ********************************************************************** DEBUG PATCH RELEASE 6.3.1D8 Based on release 6.3.1 ********************************************************************** This is a DEBUG patch, which means it has fixes that have not been released in a regular software version yet. These fixes, if validated, will be checked in and released in a future release after adequate QA and testing. Fixes contained in this patch release are: 85294 Ineffective nlm GRANTED requested (callback) 78177 Renaming a small number of files repeatedly causes WAFL dir size to grow 83697 CPU at 100% after 6.2.1 upgrade from 6.1.3R1, with low NFS Ops and Disk Ops 80286 Problem accessing ~snapshot directory when using the "nosymlink_strict_security" share option 82176 File access can fail when using a widelink that maps to a share that supports widelinks 78570 Slow FTP performance with clustering enabled. 80332 SNMP may stop responding after accessing miscGlobalStatus on FAS9xx appliances. 81905 kma.c crash in wafl_zombie_growfile_handler 83681 CIFS file owner may be set incorrectly 85651 CIFS SETUP fails with "DSA is unwilling to perform" 85653 can't set file attributes when restoring with Veritas NetBackup 81631 isp2100 not handling port_logged_out status correctly for CT requests 86862 Master filer panic during takeover - memory leak from dup_user_field() 86464 Volume SnapMirror from an F840,F880,F9XX to a platform other than F840,F880 or F9XX on restarting may lead to corruption on destination. -----Original Message----- From: Hill, Aaron [mailto:aaron.hill@navitaire.com] Sent: Sunday, February 16, 2003 2:14 PM To: 'Chris Blackmor' Cc: 'Paul McGuinness (ext. 771)'; 'toasters(a)mathworks.com' Subject: RE: Data ONTAP Upgrade Hey there, I welcome questions. Unfortunately, my answers may not be sufficient to satisfy your mental hunger. To be honest, I don't know exactly what the D8 does. I was informed that it fixes the snapmirror code. The error was found to occur only in our "specific" circumstance. So what was/is our circumstance? 1) F880 source snapmirroring 3 batch area volumes to an F820 target across a 20Mbit Ethernet WAN. 2) The amount of data change per day is high. Approximately 10-40Gb per volume. 3) We were experiencing poor throughput across the WAN link (carrier issue) and had to frequently abort the snapmirror and delete snapshots from the F880 source in order to prevent snapshot space over-running production data space. This throughput has since been addressed (before D8). 4) The F820 target panicked a couple of times and dumped to the ok prompt. One of the target volumes was found to be corrupt. It was difficult to say if the source filer caused it or the target filer developed the corruption independently. We had some cases open and ran wafl checks and scans on both filers as a result. No errors were turned up. 5) The crash dump file was ftp'd to NetApp. This lead to the 6.3.1D8 debug release. When I get some more time, I might read into it a bit further. For now, I am more than happy that they have addressed the issue and fixed it very quickly. We have not had any problems with 6.3.1 or the D8 releases. Thanks, Aaron -----Original Message----- From: Chris Blackmor [mailto:chris.blackmor@amd.com] Sent: Sunday, February 16, 2003 6:48 AM To: Hill, Aaron Cc: 'Paul McGuinness (ext. 771)'; 'toasters(a)mathworks.com' Subject: Re: Data ONTAP Upgrade What does the D8 fix for snapmirror? Inquiring minds want to know. ;) C- On Fri, Feb 14, 2003 at 07:18:19PM -0600, Hill, Aaron wrote: > We are running 6.3.1 on one filer (F880) and 6.3.1D8 (debug release for a > snapmirror problem we were having) on another (F820). 2 others still running > 6.2.2 I believe. > > As far as I know, all new F960s being shipped have 6.3.1 on them. > > So overall, I have to say I recommend it. If you are snapmirroring, make > sure you put 6.3.1D8 on your destination filer, that is the advice from our > NetApp representative. > > Thanks, > Aaron Hill > Systems Engineer > www.navitaire.com > > > -----Original Message----- > From: Paul McGuinness (ext. 771) [mailto:Paul.McGuinness@fineos.com] > Sent: Saturday, February 15, 2003 2:58 AM > To: 'toasters(a)mathworks.com' > Subject: Data ONTAP Upgrade > > Hi, > I have a F720 currently running on Data ONTAP 6.1.2R2 and everything is > running fine. > My question is to those of you 6.6.2 (the current GA release) and those of > you on 6.3.1. > > I was going to upgrade to 6.2.2, just to keep my system up to date, but > according to the NOW site there are a good few more active systems running > 6.3.1. > > Would 6.3.1 be a better upgrade to go to? I also see that there is now a 6.4 > first customer shipment release (12th Feb) but I think I'll leave that for > those of you who have filers to play with.... > :o) > > So 6.6.2 or 6.3.1? > > > > > > ************************************************************************** > The information contained in this e-mail is confidential, > may be privileged and is intended only for the use of the > recipient named above. If you are not the intended > recipient or a representative of the intended recipient, > you have received this e-mail in error and must not copy, > use or disclose the contents of this email to anybody > else. If you have received this e-mail in error, please > notify the sender immediately by return e-mail and > permanently delete the copy you received. This email has > been swept for computer viruses. However, you should > carry out your own virus checks. > > > Registered in Ireland, No. 205721. http://www.FINEOS.com > ************************************************************************** -- -------------------------------------------------------------------- * Chris Blackmor _______ | * * Advanced Micro Devices \____ | | Toto... I've * * Phone: (512) 602-1608 /| | | | a feeling we're * * Fax: (512) 602-5155 | |___| | | not in Kansas * * Email: chris.blackmor(a)amd.com |____/ \| | anymore. * -------------------------------------------------------------------- * My comments are mine, and mine alone. * --------------------------------------------------------------------

2 1

RE: SNMP values keeps changing after every reboot. any solution to fix this?
by Hackworth, Brian 31 Oct '03

31 Oct '03

Almost. DataFabric Manager uses the file system name as its key, rather than the SNMP OID index. I'm not aware of any way to trick MRTG into doing it that way. - Brian Hackworth When asked how Data Fabric Manager can handle this, the answer was that since it relies on a database, DFM is clever enough to update the index to check.

2 1

Re: SNMP values keeps changing after every reboot. any solution to fix this?
by Michel GELDENHUYS 31 Oct '03

31 Oct '03

Hi StoRage, I discussed this few weeks ago with a NetApp engineer. Unfortunately, this is a "by-design" behaviour: MIB values are loaded during boot and in a random fashion so there is no way to forecast (or enforce) what MIB index will be for a particular volume. So it seems that there is no other way than manually modifying MRTG configuration file to take the changes into account. When asked how Data Fabric Manager can handle this, the answer was that since it relies on a database, DFM is clever enough to update the index to check. I hope this will help. Kind regards, Michel Geldenhuys Danone Belgium SA/NV +32 2 776 68 23 michel.geldenhuys(a)danone.com Sto Rage© <Net_Backer(a)hotmail.com> Sent by: owner-toasters(a)mathworks.com 10/31/2003 03:42 AM To: <toasters(a)mathworks.com> cc: Subject: SNMP values keeps changing after every reboot. any solution to fix this? Every time the filer is rebooted. the snmp values for disk changes. This affects our mrtg monitoring. Is there a script that helps overcome this issue? How are others monitoring disk usage using mrtg? TIA After last reboot: NETWORK-APPLIANCE-MIB::dfFileSys.1 = STRING: "/vol/vol1/" NETWORK-APPLIANCE-MIB::dfFileSys.2 = STRING: "/vol/vol1/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.3 = STRING: "/vol/vol7/" NETWORK-APPLIANCE-MIB::dfFileSys.4 = STRING: "/vol/vol7/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.5 = STRING: "/vol/vol6/" NETWORK-APPLIANCE-MIB::dfFileSys.6 = STRING: "/vol/vol6/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.7 = STRING: "/vol/vol2/" NETWORK-APPLIANCE-MIB::dfFileSys.8 = STRING: "/vol/vol2/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.9 = STRING: "/vol/vol0/" NETWORK-APPLIANCE-MIB::dfFileSys.10 = STRING: "/vol/vol0/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.11 = STRING: "/vol/vol5/" NETWORK-APPLIANCE-MIB::dfFileSys.12 = STRING: "/vol/vol5/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.13 = STRING: "/vol/vol4/" NETWORK-APPLIANCE-MIB::dfFileSys.14 = STRING: "/vol/vol4/.snapshot" NETWORK-APPLIANCE-MIB::dfFileSys.15 = STRING: "/vol/vol3/" NETWORK-APPLIANCE-MIB::dfFileSys.16 = STRING: "/vol/vol3/.snapshot" Before the reboot: #NETWORK-APPLIANCE-MIB::dfFileSys.1 = STRING: "/vol/vol2/" #NETWORK-APPLIANCE-MIB::dfFileSys.2 = STRING: "/vol/vol2/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.3 = STRING: "/vol/vol1/" #NETWORK-APPLIANCE-MIB::dfFileSys.4 = STRING: "/vol/vol1/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.5 = STRING: "/vol/vol4/" #NETWORK-APPLIANCE-MIB::dfFileSys.6 = STRING: "/vol/vol4/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.7 = STRING: "/vol/vol6/" #NETWORK-APPLIANCE-MIB::dfFileSys.8 = STRING: "/vol/vol6/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.9 = STRING: "/vol/vol0/" #NETWORK-APPLIANCE-MIB::dfFileSys.10 = STRING: "/vol/vol0/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.11 = STRING: "/vol/vol5/" #NETWORK-APPLIANCE-MIB::dfFileSys.12 = STRING: "/vol/vol5/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.13 = STRING: "/vol/vol3/" #NETWORK-APPLIANCE-MIB::dfFileSys.14 = STRING: "/vol/vol3/.snapshot" #NETWORK-APPLIANCE-MIB::dfFileSys.15 = STRING: "/vol/vol7/" #NETWORK-APPLIANCE-MIB::dfFileSys.16 = STRING: "/vol/vol7/.snapshot"

3 2

scsi timeouts
by Bui, Marcus 31 Oct '03

31 Oct '03

I have an F820, 6.4.2P6, with dual port FC card that's getting a lot of scsi timeout msgs on the 7b port. The scsi timeouts spans all disk shelves on the 7b loop. All other loop is fine. These started when I moved the 18GB FC9 shelves from a F880c to the F820. Working the tech support, I replaced a couple of the LRC, replaced two shelf cable interconect, recheck all the cables, replaced FC card slot 7, remove/reinsert all the drives, and ran fcal check which passed all fcal tests. The core file result indicated to bug #98110 but upgraded to v6.4.2P6 did not resolve these error msgs. The filer panic a few times before running v6.4.1 that's why tech support recommends upgraded to v6.4.2P6. Right now I disconnected the second loop 7b. Any ideas? Slot: 7b Description: Fibre Channel Host Adapter 7b (Dual-channel, QLogic 2200A rev. 5) Firmware Rev: 2.1.32 PCI Bus Width: 64-bit PCI Clock Speed: 33 MHz Error Msg:------------------------------------------------------------------------ ------------------------------- Sun Oct 26 09:05:30 CST [scsi.cmd.retrySuccess:info]: Device 7b.20: request successful after retry #1: cdb 0x2f:00182000:0400. Sun Oct 26 09:05:59 CST [ispfc_timeout_4:warning]: 7b.27 (0x0400001b) (0x039a7320,0x2f:00196800:0400,0/0,1157/264/0,19642/0): command timeout, quiescing drive to allow outstanding I/O to complete. ---------------------------------------------------------------------------- ------------------------------------------- thanks, Marcus Bui

1 0

unix group in NIS and CIFS
by Stephane Bentebba 31 Oct '03

31 Oct '03

Hi all, i have a Filer running NIS and CIFS and specifically, specifying some Unix group (accessed by NIS) inside the CIFS configuration the problem is the following : the Giga card make some times to be up (as the spanning tree in the cisco switch requires 30 sec to come up) in this laps of time - NIS is unavailable (as the Filer complain at the reboot) - CIFS starts but doesn't correctly you can see the related bug 32129 : http://now.netapp.com/AskNOW/highlight_html.jsp?url=http%3A%2F%2Fnow.netapp… my question is : does anyone already encountered and solved this problem in another way that Netapp say (forgetting about Unix group in CIFS config / copy the NIS group definition locally on the Filer) the costumer is thinking about desengage the spanning tree at the switch side , i am not sure it will succeed -- jetez un oeil ici http://www.sensiva.com/ -- -- have a look here http://www.sensiva.com/ --

2 1

msdos 8.3 name and solstice
by Stephane Bentebba 31 Oct '03

31 Oct '03

Hi all, - one of our Filer is running with NFS license and NIS - some Windoz client are using Solstice to connect on the Filer Solstice = a pcnfsd client that connect to a Unix machine with the pcnfsd server - this way, all pcnfsd Windows client can connect on the NFS Filer the problem is really strange not all but some only few of them encounter this problem > when copying files with long names to the Filer (throught Solstice), the filename is changed to a short version : "my document.doc" tunr into "mydocu~1.doc" on the Filer - fro; the Windoz client point of view from a Unix machine, the file remains with a long normal name "my document.doc" when hitting the Up button then going in the directory again, the file appear normally again (long form) this strange behavior implies numerous of problems does anyone already encounter this problem ? the Filer used to be in 537r3, upgrading the Filer in 613r2 doesn't change the problem ps : solstice was a Sun product that is no more supported... -- -- jetez un oeil ici http://www.sensiva.com/ -- -- have a look here http://www.sensiva.com/ --

1 0

ndmpcopy loopback perf
by George Kahler 31 Oct '03

31 Oct '03

Hi, when I ndmpcopy from say, volA to volB on the same head does ndmpcopy use the loopback to do this ? If so, at what speed can the loopback run at ? I can not seem to go faster then about 20 GB/hr (about 50 Mbits/sec). I do have GigE NICs in the filer for regular nfs connections to clients. 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

5 5

AIX write performance
by bill_link＠dstoutput.com 29 Oct '03

29 Oct '03

Toasters, I am having difficulty with large file writes to an f720 from any of my IBM hosts, if I start a large write with dd then try to do anything on the filer from that host it hangs until the write completes. From a different host access to the filer is near normal. I do not have this problem with linux boxes on the same network so it seems that the problem lies with AIX somehow. IBM has been sympathetic although not very helpful. Any thoughts would be much appreciated. Bill Link System Admin Dstoutput inc.

6 6

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 October 2003