toasters

toasters@lists.teaparty.net

13514 discussions

Spurious tape drive errors in 6.1+?
by Jeffrey Krueger 28 Aug '01

28 Aug '01

Has anyone else noticed *lots* of these messages? Mon Aug 27 15:00:53 PDT [fuelcell: Java Thread:warning]: Ndmpd8: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:00:53 PDT [fuelcell: Java Thread:warning]: Ndmpd9: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:03:58 PDT [fuelcell: Java Thread:warning]: Ndmpd10: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:03:58 PDT [fuelcell: Java Thread:warning]: Ndmpd11: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:07:02 PDT [fuelcell: Java Thread:warning]: Ndmpd12: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:07:03 PDT [fuelcell: Java Thread:warning]: Ndmpd13: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:10:07 PDT [fuelcell: Java Thread:warning]: Ndmpd14: IOException: Device I/O error. Device may have no tape Mon Aug 27 15:10:08 PDT [fuelcell: Java Thread:warning]: Ndmpd15: IOException: Device I/O error. Device may have no tape We've got two DLT7000's hooked up as nrst0a and nrst1a to an F840 - this seems to happen for both drives. 6.1.1 loves telling us this error. We seem to be able to get backups (except for the bug on two of our volumes). This is occurring on two F840's, two others work fine. Do I just need to fix my tape drives or is anyone else noticing this problem as well? -- Jeff

2 3

Backup Software
by Barry Lustig 27 Aug '01

27 Aug '01

Is there a general consensus on backup software for the files? The choices for backing up via NDMP seems to be growing and I'm curious what others are using. I've started looking at syncsort's product, bakbone's netvault, veritas, and time navigator. Each of the packages so far seems to have a pretty good mix of good and bad points. barry

1 0

Re: FC Disk Replacement
by jeff.mery＠ni.com 27 Aug '01

27 Aug '01

The filer will always try to use the smallest disk possible to replace a failed disk. In your case it will grab the 36GB drive. It will only use larger drives if they are the only spares on the system. Jeff Sateesh Mucharla <sateesh(a)ampere.ns To: toasters(a)mathworks.com c.com> cc: Sent by: Subject: FC Disk Replacement owner-toasters@mat hworks.com 08/27/2001 04:57 PM Please respond to Sateesh Mucharla Hi All, I am curious to know what to do in mixed FC environment when a disk fails. I have 36GB and 72GB disks (all are FC) on a F760. I have volumes with 36GB drives also a seperate volume with 72GB drives. I have one 36GB and one 72GB drive as hot spares. Now one of my 36GB FC drive is logging errors excessively. I want to replace this disk with a spare one. I believe, i have to do "disk fail <diskID>" to fail the erring disk so that it can pick up one that is a hot spare. My question/worry is that i do not want the 72GB disk to be replaced for a 36GB. How can i ensure that only 36GB drive is picked up for a 36GB drive that is failed. Thanks and regards. Sateesh M National Semiconductor Corp.

1 0

FC Disk Replacement
by Sateesh Mucharla 27 Aug '01

27 Aug '01

Hi All, I am curious to know what to do in mixed FC environment when a disk fails. I have 36GB and 72GB disks (all are FC) on a F760. I have volumes with 36GB drives also a seperate volume with 72GB drives. I have one 36GB and one 72GB drive as hot spares. Now one of my 36GB FC drive is logging errors excessively. I want to replace this disk with a spare one. I believe, i have to do "disk fail <diskID>" to fail the erring disk so that it can pick up one that is a hot spare. My question/worry is that i do not want the 72GB disk to be replaced for a 36GB. How can i ensure that only 36GB drive is picked up for a 36GB drive that is failed. Thanks and regards. Sateesh M National Semiconductor Corp.

1 0

RE: Uniquely identifying a cluster, and cluster behavior
by Fox, Adam 27 Aug '01

27 Aug '01

If you're willing to do some SNMP queries, there are some pretty good ones. Pay particular attention to cfSettings, cfPartnerName, cfPartnerSysid, etc. BTW: This comes from /etc/mib/netapp.mib on your filer. -- Adam Fox NetApp Professional Services, NC adamfox(a)netapp.com -- the Clustered Failover group cfSettings OBJECT-TYPE SYNTAX INTEGER { notConfigured(1), enabled(2), disabled(3), takeoverByPartnerDisabled(4), thisNodeDead(5) } ACCESS read-only STATUS mandatory DESCRIPTION "Clustered Failover configuration settings. The value notConfigured(1) indicates that the cluster is not licensed. The thisNodeDead(5) setting indicates that this node has been taken over." ::= { cf 1 } cfState OBJECT-TYPE SYNTAX INTEGER { dead(1), canTakeover(2), cannotTakeover(3), takeover(4) } ACCESS read-only STATUS optional DESCRIPTION "Clustered Failover state." ::= { cf 2 } cfCannotTakeoverCause OBJECT-TYPE SYNTAX INTEGER { ok(1), unknownReason(2), disabledByOperator(3), interconnectOffline(4), disabledByPartner(5), takeoverFailed(6) } ACCESS read-only STATUS optional DESCRIPTION "The reason this node cannot takeover its partner" ::= { cf 3 } cfPartnerStatus OBJECT-TYPE SYNTAX INTEGER { maybeDown(1), ok(2), dead(3) } ACCESS read-only STATUS optional DESCRIPTION "This is the status last returned by the Clustered Failover partner." ::= { cf 4 } cfPartnerLastStatusUpdate OBJECT-TYPE SYNTAX TimeTicks ACCESS read-only STATUS optional DESCRIPTION "If non-zero, the absolute time of the last update from the Clustered Failover partner." ::= { cf 5 } cfPartnerName OBJECT-TYPE SYNTAX DisplayString ACCESS read-only STATUS optional DESCRIPTION "Hostname for Clustered Failover partner." ::= { cf 6 } cfPartnerSysid OBJECT-TYPE SYNTAX INTEGER ACCESS read-only STATUS optional DESCRIPTION "System ID for Clustered Failover partner." ::= { cf 7 } cfInterconnectStatus OBJECT-TYPE SYNTAX INTEGER { notPresent(1), down(2), partialFailure(3), up(4) } ACCESS read-only STATUS optional DESCRIPTION "The current status of the interconnect. The partialFailure(3) status indicates that one of the links in the interconnect hardware is down." ::= { cf 8 } > -----Original Message----- > From: Julius Talbot [mailto:jstalbot@mail.com] > Sent: Monday, August 27, 2001 4:56 PM > To: toasters(a)mathworks.com > Subject: Uniquely identifying a cluster, and cluster behavior > > > How does one uniquely identify a cluster? > > I am trying to write some utilities that will tell me > filesystem usage and the like, and would like to uniquely > identify filers and clusters. I could use the system ids, but > wouldn't it be impossible to determine the system id of a > failed partner? > > Ideally, I want a single text string that identifies the pair > as a whole. > > Anyone have an idea? > -- > > _______________________________________________ > FREE Personalized E-mail at Mail.com > http://www.mail.com/?sr=signup > > Talk More, Pay Less with Net2Phone Direct(R), up to 1500 > minutes free! > http://www.net2phone.com/cgi-bin/link.cgi?143 >

1 0

Re: SnapMirror to tape?
by Steven Boyer 27 Aug '01

27 Aug '01

I have been experimenting with SnapMirror to tape mostly for initial mirroring to a remote site. As far as I know there is no way to catalog a SnapMirror tape, and you would have to restore the whole volume not just a missing or corrupted file. Also I believe you will need ONTAP 6.1 for your configuration to work. Steven ---------- Original Message ---------------------------------- From: "Gifford, Josh" <Josh.Gifford(a)ptd.siemens.com> Date: Mon, 27 Aug 2001 11:33:29 -0400 >We have implemented a SnapMirror strategy where a small site in Atlanta is >mirroring to our larger filer in our data center in Raleigh. We are doing >this in place of tape backups because there is no local IT staff to switch >out tapes and keep an eye on things. The mirror seems to be working fine, >but I had hoped to be able to send the snap mirrored volume to tape once a >week so that they could be sent off site. The problem is that the >destination volume is listed as "snapmirrored," not "online" like the rest. >Veritas will back up all of the other volumes, but not the snapmirrored one. > >It's not a big deal because I have all of the data in two places, but it >would be nice to also have a current copy of the data an underground in a >vault. Let me know what you guys think my options might be or what you guys >have done in these situations. Thanks! > >Best Regards, > >Josh J Gifford MCP >Senior NT Administrator >Siemens Power Transmission & Distribution >7000 Siemens Rd >Wendell, NC 27591 >Phone: (919) 365-2806; Fax: (919) 365-1080 >Email: <mailto:josh.gifford@ptd.siemens.com> josh.gifford(a)ptd.siemens.com > > > > > -- Steven N. Boyer UNIX/NT/NetApp Support HQUSCENTCOM General Dynamics --

1 0

RE: SnapMirror to tape?
by Knight, Jason 27 Aug '01

27 Aug '01

You can backup a snapmirrored volume, but you need to use a snapshot that already exists in the volume everytime you do the backup, because you cannot create a "snapshot_for_backup" in a read only volume. We have a similar situation and we use hourly.0 to backup the snapmirrored volumes. Jason Knight -----Original Message----- From: Gifford, Josh [mailto:Josh.Gifford@ptd.siemens.com] Sent: Monday, August 27, 2001 10:33 AM To: 'Toasters discussion list (E-mail)' Subject: SnapMirror to tape? We have implemented a SnapMirror strategy where a small site in Atlanta is mirroring to our larger filer in our data center in Raleigh. We are doing this in place of tape backups because there is no local IT staff to switch out tapes and keep an eye on things. The mirror seems to be working fine, but I had hoped to be able to send the snap mirrored volume to tape once a week so that they could be sent off site. The problem is that the destination volume is listed as "snapmirrored," not "online" like the rest. Veritas will back up all of the other volumes, but not the snapmirrored one. It's not a big deal because I have all of the data in two places, but it would be nice to also have a current copy of the data an underground in a vault. Let me know what you guys think my options might be or what you guys have done in these situations. Thanks! Best Regards, Josh J Gifford MCP Senior NT Administrator Siemens Power Transmission & Distribution 7000 Siemens Rd Wendell, NC 27591 Phone: (919) 365-2806; Fax: (919) 365-1080 Email: <mailto:josh.gifford@ptd.siemens.com> josh.gifford(a)ptd.siemens.com

1 0

SnapMirror to tape?
by Gifford, Josh 27 Aug '01

27 Aug '01

We have implemented a SnapMirror strategy where a small site in Atlanta is mirroring to our larger filer in our data center in Raleigh. We are doing this in place of tape backups because there is no local IT staff to switch out tapes and keep an eye on things. The mirror seems to be working fine, but I had hoped to be able to send the snap mirrored volume to tape once a week so that they could be sent off site. The problem is that the destination volume is listed as "snapmirrored," not "online" like the rest. Veritas will back up all of the other volumes, but not the snapmirrored one. It's not a big deal because I have all of the data in two places, but it would be nice to also have a current copy of the data an underground in a vault. Let me know what you guys think my options might be or what you guys have done in these situations. Thanks! Best Regards, Josh J Gifford MCP Senior NT Administrator Siemens Power Transmission & Distribution 7000 Siemens Rd Wendell, NC 27591 Phone: (919) 365-2806; Fax: (919) 365-1080 Email: <mailto:josh.gifford@ptd.siemens.com> josh.gifford(a)ptd.siemens.com

1 0

RE: Filer hung up tight
by William Liu 24 Aug '01

24 Aug '01

Same thing happened to me 2 days ago on a F740 filer. Last thing on LED was 28 ops/sec. Couldn't get to it whatsoever. Had to reboot by shutting down power supplies. Filer had been up over 320+ days. I called NetApp and they said they could do nothing since there was no core dump. There was only a "bounds" file with "0 0" in it. They told me if it happens next time to call them and do the LED reset for a coredump. -----Original Message----- From: Maglinger, Paul [mailto:PMAGLINGER@scvl.com] Sent: Friday, August 24, 2001 6:30 AM To: 'Sam Rafter' Cc: 'NetApps list server' Subject: RE: Filer hung up tight I checked with my boss, who got in before I did and he said that he tried filer view after trying the console. It came back as "Page not found". -----Original Message----- From: Sam Rafter [mailto:rafter@netapp.com] Sent: Thursday, August 23, 2001 18:49 To: Maglinger, Paul Cc: 'NetApps list server' Subject: Re: Filer hung up tight You do have one alternative in this situation. There's a synch button on the front of your filer. On the 700 series, it's the little hole under the LEDs. When you press that, the filer should drop to the ok prompt. From there, you can type 'sync'. There's a NOW article on doing just very thing: http://now.netapp.com/Knowledgebase/solutionarea.asp?id=3.0.620867.2715498&r esource= In the 'nothing was working' checklist, was NFS/CIFS and FilerView checked as well? 850 ops is a lot to do nothing at all. If the number never fluctuated, I'd have my suspicions that it wasn't getting updated. We have watchdog systems in place to catch such things, so I'm a bit mystified. -Sam * It was Thu, Aug 23, 2001 at 04:26:24PM -0500 when Maglinger, Paul wrote: > > We have a 740 filer running 6.0.1R3. Last night I got a page that > "nothing was working". I headed into work and found that: > > > > 1) I could not telnet into the filer. > > 2) I could not get in through console either. > > 3) The display on the front of the filer said 850 ops but the activity > led wasn't flashing at all. > > > > I figured I had no choice, so I shut it down hard and brought it back > up. Everything came up fine. The message file posted a time stamp at > 11:00pm but nothing at midnight (this was 1:30am). The message file > also said that the fiiler was down for 6057 secs. The crash folder > had one file in it called "bounds" and just contained "-1 -1". > Backups were being performed at the time using Veritas Backup Exec. > The backup log said everything was going fine until it found a > "corrupted file", then it started skipping files until the filer was > rebooted and it started in backing up again. We're trying to figure > out what went wrong but didn't have much to go on. Anyone have any > thoughts where to look now (or if this happens again)? > > > > "No matter where you go, there you are." > > > > Paul Maglinger > > System Administrator > > Shoe Carnival, Inc. --

1 0

Concurrent File Write Access on Filer CIFS
by Jay Yau 24 Aug '01

24 Aug '01

For Filer CIFS, when concurrently several clients open the same file, then thees clients (all these clients have write access on the file) try to modify and write to this file differently almost at the same time. What will happen to this file? Which client will have the final right to have the file modified? Thanks for help! -- Jay Yau Automated Systems (HK) Ltd. Direct: (852) 26083741 Fax: (852) 26086589 Email: jayau(a)asl.com.hk

3 2

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