toasters August 2001

toasters@lists.teaparty.net

122 participants
161 discussions

RE: migrating data from volume to volume
by Clawson, Simon 28 Aug '01

28 Aug '01

NDMP Copy seems to work really well for this type of job. Just remeber to put /vol/volx/ before your target path. I forgot and created a huge file in my root volume! Simon > -----Original Message----- > From: White, Lance [mailto:Lance.White@netapp.com] > Sent: Tuesday August 2001 14:32 > To: 'Lance Murayama'; toasters(a)mathworks.com > Subject: RE: migrating data from volume to volume > > > NDMPcopy or 'vol copy' would do this. > > -----Original Message----- > From: Lance Murayama [mailto:opi@cognigine.com] > Sent: Tuesday, August 28, 2001 6:22 AM > To: toasters(a)mathworks.com > Subject: migrating data from volume to volume > > > I am about to migrate 100GBytes of data from vol0 to a new > vol1 that I just added > to my F740. In the past I have always used tar | tar or dump > | restore. I have not done > this in quite some time. Are there better/faster ways to do > this? Any suggestions or > tips would be greatly appreciated. > > Also the 100GB that is being moved is a single qtree. I don't > know if that makes a > difference but I thought I would mention it. Thanx... >

1 0

RE: Backup Software
by Jay Newton 28 Aug '01

28 Aug '01

My organization is using CommVault's Galaxy. It narrowed down to that or SyncSort. If you are and NT shop with Exchange, CommVault is the way to go from a price/feature perspective. We have thus far been very pleased with our choice. We get very impressive throughput in NDMP both with direct attached devices and across the network. http://www.commvault.com -----Original Message----- From: Barry Lustig [mailto:barry@lustig.com] Sent: Monday, August 27, 2001 5:52 PM To: Toasters Subject: Backup Software 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

Uniquely identifying a cluster, and cluster behavior
by Julius Talbot 28 Aug '01

28 Aug '01

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

2 1

RE: migrating data from volume to volume
by White, Lance 28 Aug '01

28 Aug '01

NDMPcopy or 'vol copy' would do this. -----Original Message----- From: Lance Murayama [mailto:opi@cognigine.com] Sent: Tuesday, August 28, 2001 6:22 AM To: toasters(a)mathworks.com Subject: migrating data from volume to volume I am about to migrate 100GBytes of data from vol0 to a new vol1 that I just added to my F740. In the past I have always used tar | tar or dump | restore. I have not done this in quite some time. Are there better/faster ways to do this? Any suggestions or tips would be greatly appreciated. Also the 100GB that is being moved is a single qtree. I don't know if that makes a difference but I thought I would mention it. Thanx...

1 0

migrating data from volume to volume
by Lance Murayama 28 Aug '01

28 Aug '01

I am about to migrate 100GBytes of data from vol0 to a new vol1 that I just added to my F740. In the past I have always used tar | tar or dump | restore. I have not done this in quite some time. Are there better/faster ways to do this? Any suggestions or tips would be greatly appreciated. Also the 100GB that is being moved is a single qtree. I don't know if that makes a difference but I thought I would mention it. Thanx...

1 0

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

← Newer
1
2
3
4
5
6
7
8
...
17
Older →

Jump to page:

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 August 2001