toasters

toasters@lists.teaparty.net

13520 discussions

netapp not writing disks to /etc/.broken_disks when they fail
by Daniel Finn 07 Nov '02

07 Nov '02

It's an F740 running ONTAP 5.3.5R2. Is this normal? All the rest of the netapps we have write to /etc/.broken_disks when a disk fails but they are also running a more up to date version of ONTAP. Is it because of the older version?

1 0

F540 + Spares available
by Chuck Tomasi 06 Nov '02

06 Nov '02

I saw someone mention they had some old equipment (F630) on here. I was wondering if anyone was interested in picking up some "spare parts" for a NetApp. We have an F540 with two sets of diff SCSI 4GB shelves and two sets of diff SCSI 9GB shelves plus lots of spares (minus the motherboard.) If anyone is interested I can provide more details one-on-one. --Chuck Tomasi --Corp. Systems Administrator --Plexus Corp.

2 1

Re: Crossroads Problems
by Rune Bakken 06 Nov '02

06 Nov '02

Paul Caron <caron(a)sig.com> writes: > The problem we are experiencing is that the devices on our filers which > represent the SAN attached tape drives (shown by sysconfig -t) get randomly > recreated every time the Crossroads reboots. This was a problem that I encountered early in ONTAP 6.0.x (and 6.1.x to?) although sporadic, often enough to write a perl script to reset the tape aliases a few times a week. The LUN did not change, but the filers decided to rename the the aliases after the fc-card was reset or unplugged from the Brocade. (The last was a precautionary action to prevent the filer from panic'ing from a SCSI reset when the bridge or library was rebooted) > It appears as though the Crossroads is not presenting the tape > drives available on the ATL P7000 in the same order every time it > reboots. Although I am using the STK (Crossroad) 3250, I would believe that the menus/commands are similiar. Have you set any values in the "Fibre Channel to SCSI mapping"/ "Mapping Table# ?? Does the values change after reboot? What about "Display Attached SCSI Devices", does this change between reboots? > Has anyone else run into this situation? Is there any way of compensating > for this on the Netapp filers? Without the Crossroads not keeping the LUN values stable, you are out of luck. As far as I know there is no way of keeping track of a changing LUN. > > BTW, we have received suggestions regarding using the Netapp "storage > alias" and "storage unalias" commands. However, since the Galaxy backup > software relies on the actual device names (i.e. rst0l, nrst0l, urst0l, > etc.) this hasn't helped us so far. The Filer actually sees the real LUN value, but you may assign an alias, such as st01 st02 ... the device names rst02,nrst02,urst02 etc gives the device functionality. To sum up: If the Crossroads keep changing the LUN values, I would suggest you bring your vendor to bring this case directly to Crossroads. >>>>>>>>>>>>>>>>>>>.rune -- Rune Bakken Senior System Administrator Telenor Business Solutions (Nextra), ManagedServices/*/IP-Services Ph: +47 992 86 474

3 2

RE: Storagetek l80
by James Beal 06 Nov '02

06 Nov '02

Hello, We just became the proud owners of an L80. Its a bit of a big beast, and has baffled me a bit! I was wondering if anyone out there has one of these attached directly to the netapp, and also attached to a backup server such as an E250? During the selection process the company we engaged told me that each drive could be connected directly to a machine via SCSI. Looking at the back of the library, I think this is do-able, as each drive has SCSI connections, but I am not sure if these are intended to be connected to the library itself, then daisy chained together, or if the drive can be attached to the filer... I am waiting on a colleague to chase the supplier, but thought I put out request to you guys too... Okay, I don't have a toaster at present but I have looked after them in the past. I do have a L80 which is a great tape library. Each of the tape drives can be connected to a different system and storage tek will say if you have LTO drives in it then each drive should really be on a separate bus for performance/reliablity issues. What software are you using to drive this, netback from veritas can handle one having a robotic with the tape drives attached to different machines, for a toaster you would need to have a NDMP license as well, the robotic being controlled by the E250.

1 0

Re: disk_erase_label
by Chris Blackmor 06 Nov '02

06 Nov '02

Under priv set diag you can us the disk unfail command to make a broken disk a spare again but this is (as you said) used at your own risk. Is that what you are looking for? I have not had to use this command myself so I don't know what the real-world implications are. C- m5*> disk usage: disk <options> Options are: fail <disk_name> - fail a file system disk remove <disk_name> - remove a spare disk swap - prepare (quiet) bus for swap unswap - undo disk swap and resume service scrub { start | stop } - start or stop disk scrubbing unfail <disk_name> - make a failed disk a spare again shm_stats [<disk_name>] - Storage Health Monitor stats for a disk On Mon, Nov 04, 2002 at 04:55:35PM -0000, Edward Hibbert wrote: > I powered off a NetApp rather abruptly. Oops. > > Since then, one of my disks reckons it's broken. > > nacolt1> sysconfig -r > Volume vol0 (root) > > RAID group 0 > > RAID Disk HA.ID HA SHELF BAY CHAN Used (MB/blks) Phys > (MB/blks) > --------- ----- ------------ ---- -------------- > -------------- > parity 8.3 8 0 3 FC:A 34500/70656000 > 35003/71687368 (scrubbing RAID group, 20% done) > data 8.4 8 0 4 FC:A 34500/70656000 > 35003/71687368 > data 8.1 8 0 1 FC:A 34500/70656000 > 35003/71687368 > data 8.0 8 0 0 FC:A 34500/70656000 > 35003/71687368 > > Volume vol1 > > RAID group 0 > > RAID Disk HA.ID HA SHELF BAY CHAN Used (MB/blks) Phys > (MB/blks) > --------- ----- ------------ ---- -------------- > -------------- > parity FAILED N/A 34500/70656000 > data 8.6 8 0 6 FC:A 34500/70656000 > 35003/71687368 > data 8.5 8 0 5 FC:A 34500/70656000 > 35003/71687368 > > > I believe there's a command disk_erase_label which I can use to (at my own > risk) try to recover this disk. But it needs a disk id parameter. I'm > guessing that the disk id is 8.2 (the missing one from the info above) - but > that doesn't work. > > nacolt1*> disk_erase_label 8.2 > bad disk_name: 8.2 > > Is there some other disk id I need to find out - or is this just an > indication that the disk really is broken? > > This is an old ONTAP. version says: > > NetApp Release /n/snapon/dnoveck/537: Tue May 29 08:30:47 PDT 2001 > > Thanks for any help, > > Edward Hibbert > Internet Applications Group > Data Connection Ltd > Tel: +44 131 662 1212 Fax: +44 131 662 1345 > Email: eh(a)dataconnection.com Web: http://www.dataconnection.com > -- ----------------------------------------------------------------------- * Chris Blackmor _______ | * * Advanced Micro Devices \____ | | My comments are * * Phone: (512) 602-1608 /| | | | mine, and mine * * Fax: (512) 602-5155 | |___| | | alone. * * Email: chris.blackmor(a)amd.com |____/ \| | * ----------------------------------------------------------------------- * "Forty-two!" he said with a malicious grin. * * "No, doesn't work. Nevermind." * * Restaurant at the End of the Universe * * Hitchhikers Guide to the Universe * -----------------------------------------------------------------------

1 0

nfs file size limits
by Daniel Finn 05 Nov '02

05 Nov '02

what are they? I came across some posts while searching groups.google.com and some people mentioned that it's 4G for nfs v2 and higher for v3. Does anyone on here know for sure? Thanks Dan

1 0

Storagetek l80
by Clawson, Simon 05 Nov '02

05 Nov '02

Hello, We just became the proud owners of an L80. Its a bit of a big beast, and has baffled me a bit! I was wondering if anyone out there has one of these attached directly to the netapp, and also attached to a backup server such as an E250? During the selection process the company we engaged told me that each drive could be connected directly to a machine via SCSI. Looking at the back of the library, I think this is do-able, as each drive has SCSI connections, but I am not sure if these are intended to be connected to the library itself, then daisy chained together, or if the drive can be attached to the filer... I am waiting on a colleague to chase the supplier, but thought I put out request to you guys too... Anyone? 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

VoIP Solution
by Mark 05 Nov '02

05 Nov '02

Customer reference: IntelFone.com VoiceMaster is the ultimate hardware/software platform to enable your VoIP infrastructure. Our solution offers great benefits at the lowest prices in the industry! VoiceMaster offers the most comprehensive solution that includes all services necessary to support pre-paid/post-paid calling card applications: IVR Server - provides voice media files play and storage functionality. Radius Server - allows PIN/Card Number authentication and authorization. Radius server supports three levels of functionality: a.. Authentication of PIN/Card Number and account balance feedback (e.g. "You have $30 dollars and 26 cents remaining"); b.. Authorization of destination phone number and calling time feedback (e.g. "You have 326 minutes remaining"); c.. Call record to provide total calling time and different parameters of the completed call; d.. Radius server supports database interface to Sybase 11.0.3 database (royalty free). All account parameters, call rates, and records are kept in the Sybase database server. Third-party applications are supported via ODBC/JDBC interfaces. Sybase 11.0.3 database server - provides robust industrial database storage and support for multiple concurrent connections with complete database backup and recovery functionality. Full VoIP schema is developed and can me modified if customer has custom specifications. Web Server - to allow user profile storage and account management. Provides facilities for detailed call history retrieval and account balance management. Web Based Credit Card interface - allows online account funding via credit cards. Also support automated account balance re-charges. VoIP Gateway/Gatekeeper (optional) - provides PSTN origination/termination services and communicates with the IVR and Radius server to provide scaleable, industrial performance. All systems are fully integrated and provides in a custom format based on the Demo Site that you can review. Customization of web graphical interface and credit card processing is available. All solution modules are pre-installed on a hardware platform (1U rack-mountable Industrial server). The customer may purchase the VoIP gateway directly from us as part of our solution or buy it from a third-party that supports IVR and Radius authentication. Mark Forst Global Account Manager Deliverus.net Sun Specialists of Value 408.896.1220 Storage Solutions Networking/IP Solutions Database Solutions www.deliverus.net

1 0

RE: Crossroads Problems
by Paul Caron 04 Nov '02

04 Nov '02

Thanks for your responses. Here are some answers to the questions that were presented. > Have you set any values in the "Fibre Channel to SCSI mapping"/ "Mapping Table#"? We have a custom map that we're using to map the individual tape drives to Crossroad LUNs. > Do the values change after reboot? No. > What about "Display Attached SCSI Devices", does this change between reboots? No. The attached SCSI devices are always the same. ---------------------------------- The "Fibre Channel to SCSI Mapping Mode" for our Crossroads is not set to auto. We are using a custom map to make sure that each tape drive always has the same LUN. We are not using the "index" map since that one is currently being used for our Windows media agent. However, there doesn't appear to be a difference between custom maps and the "index" map other than the name of the "index" map cannot be changed. If I am mistaken, please let me know. Also, I guess I need to qualify my details a bit more. The LUNs presented by the Crossroads are fixed in that they are always refer to the same drive on the P7000. The problem is that the LUNs get assigned to different devices on the filer. For example, LUN 2 will be assigned to the rst1l, nrst1l, ..., urst1a devices after one of the Crossroads reboots. After another Crossroads reboot, LUN 3 will be assigned to those same devices on the filer. We need to preserve a consistent device mapping on the filer. Reassigning storage aliases seems to be the best workaround we've found so far. My quick hack script to do this is below (the x:xxx:xxxxxx:xxxxxx represents the WWN of the Crossroads). We're still looking for a better long term approach. Either the Crossroads needs to present the LUNs in the same order every time, or the Netapp filer needs to have a way to specify a persistent device-to-LUN mapping for FCAL attached devices. ----------------------------------------------------------------------- #!/bin/sh FILER=$1 if [ -z "$FILER" ]; then echo "usage: resetaliases <filername>" exit 1 fi /bin/rsh $FILER storage unalias -a /bin/rsh $FILER storage alias mc0 WWN[x:xxx:xxxxxx:xxxxxx] /bin/rsh $FILER storage alias st0 WWN[x:xxx:xxxxxx:xxxxxx]L1 /bin/rsh $FILER storage alias st1 WWN[x:xxx:xxxxxx:xxxxxx]L2 /bin/rsh $FILER storage alias st2 WWN[x:xxx:xxxxxx:xxxxxx]L3 /bin/rsh $FILER storage alias st3 WWN[x:xxx:xxxxxx:xxxxxx]L4 /bin/rsh $FILER storage alias st4 WWN[x:xxx:xxxxxx:xxxxxx]L5 /bin/rsh $FILER storage alias st5 WWN[x:xxx:xxxxxx:xxxxxx]L6 /bin/rsh $FILER storage alias st6 WWN[x:xxx:xxxxxx:xxxxxx]L7 /bin/rsh $FILER storage alias st7 WWN[x:xxx:xxxxxx:xxxxxx]L8 exit 0 -----------------------------------------------------------------------

1 0

RE: Crossroads Problems
by Mike Ball 01 Nov '02

01 Nov '02

Paul, It sounds like you are running the Crossroads router in "Fibre Channel to SCSI Mapping Mode" of auto. Change the mode to "Indexed" and this should fix your problem. The mapping mode determines how the targets behind the router (i.e. tape that are attached to the SCSI side of the router) are seen by Fibre Channel hosts on the network. Specifically, the various addressing modes determine how Fibre Channel LUNs are mapped to the SCSI peripherals on the router, and how that mapping can change when you reboot or reconfigure the router. You normally want to use indexed mode to ensure that particular target/luns on the SCSI side of the router are persistently presented as the same LUNs on the Fibre Channel side of the router, regardless of whether devices have been removed on the SCSI side or were turned off when the router booted. Mike -----Original Message----- From: Paul Caron [mailto:caron@sig.com] Sent: Friday, November 01, 2002 11:02 AM To: toasters(a)mathworks.com Subject: Crossroads Problems We are currently trying to implement a SAN backup solution involving several F840 (ONTAP 6.2R2, latest FW) and R100 (ONTAP 6.2.1, latest FW) filers, a Brocade Silkwork 3800 switch (kernel 5.3.1, fabric OS 3.0), a Crossroads 10000, an ATL P7000 tape drive library with 8 SDLT drives, and Commvault Galaxy software (4.1.0, Build 26). We have tried firmware versions 4.03.16, 4.03.17, and 4.19 on the Crossroads 10000. The problem we are experiencing is that the devices on our filers which represent the SAN attached tape drives (shown by sysconfig -t) get randomly recreated every time the Crossroads reboots. Since we can't guarantee which devices will get assigned to which tape drive LUN, they often do not match those configured in the Commvault Galaxy backup software. The resulting mismatch causes many backup problems. It appears as though the Crossroads is not presenting the tape drives available on the ATL P7000 in the same order every time it reboots. We have read through the information available at http://www.netapp.com/osn/info/config.html#fcsan. This site shows support for the Brocade 3800. Although they do not call out support for the Crossroads 10000 specifically, they do show a wide range of options using various SCSI-FIBER bridges with a variety of libraries. The StorageTek library they show uses a Crossroads 42xx series switch. Has anyone else run into this situation? Is there any way of compensating for this on the Netapp filers? I realize that it would be better if the Crossroads would present the tape drive LUNs in the same order every time, but that doesn't appear to be configurable. It would help if we could permanently assign the devices to the tape drive LUNs on the filers. BTW, we have received suggestions regarding using the Netapp "storage alias" and "storage unalias" commands. However, since the Galaxy backup software relies on the actual device names (i.e. rst0l, nrst0l, urst0l, etc.) this hasn't helped us so far.

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