toasters

toasters@lists.teaparty.net

13531 discussions

NDMPCOPY solved
by Haber, David J. 18 Dec '02

18 Dec '02

Let me thank everyone one more time. I was successful in copying file from the 840 filer (6.1.1R2) to the NearStore (6.2.1) without upgrading the 840. By running this command from the NearStore it worked: NEARSTORE> ndmpcopy -sa root:rootpassword -da root:rootpassword -st text -dt text 840:/vol/vol0/Test NearStore:/vol/vol0/test Thanks to everyone's input I realized I needed to add the -st option and, after awhile the obvious was revealed---the path needs is case sensitive. Once I had these two elements in place everything worked. So thanks again for everyone's help. I also now have fresh, new root passwords for all of my filers!!!

1 0

RE: Ndmpcopy
by Chitre, Amol 18 Dec '02

18 Dec '02

Hi David, Let me clear some of the confusion for you.... ndmpcopy command on the filer acts like an application. It was introduced in Ontap 6.2. Hence you need to have the filer running the ndmpcopy command at Ontap 6.2 or higher. You can run the ndmpcopy command on any one of the following filers: 1. the source filer for the copy [ source has to be atleast 6.2, destination atleast 5.3.7 ] 2. destination filer for the copy [ destination has to be atleast 6.2, source atleast 5.3.7 ] 3. filer which is neither the source nor destination. [ filer has to be atleast 6.2, source and destination atleast 5.3.7 ] Basically it makes 2 ndmp connections to the ndmp servers on the source filer and the destination filer. The source and destination must support ndmp version 3 or higher. Now, looking at your last email, I see you have waded through the initial issues. Currently, there seems to be trouble kicking off the dump. Check if the path supplied on the ndmpcopy command line is correct, and it exists. So from the command line, it seems like you are running the ndmpcopy command on a machine NEARSTORE. And you are copying (source) alifiler:/vol/vol0/test to (destination)pacsnas1:/vol/vol0/test. Let me know if this is what you intended. If so, please send me the ndmpd debug logs (level 50). Thanks, Amol. *-----Original Message----- *From: Haber, David J. [mailto:David.Haber@vtmednet.org] *Sent: Wednesday, December 18, 2002 10:31 AM *To: 'toasters(a)mathworks.com' *Subject: Ndmpcopy * * *Thanks to everyone for their responses. I am getting closer to *a solution. *The consensus from the NetApp engineers is that I need to *upgrade the source *filer (840) to at least version 6.2.1. Others have told me *that I should be *able to run ndmpcopy from a third machine and still be able to copy the *files from the 840 to the NearStore even though the 840 is not *at 6.2.1. *This is where I'm at before the upgrade: * *NEARSTORE> ndmpcopy -sa root:dedsam9ap -da root:dedsam9ap -st *text -dt text *alifiler:/vol/vol0/test p *acsnas1:/vol/vol0/test *Ndmpcopy: Starting copy [ 21 ] ... *Ndmpcopy: alifiler: Connect: Connection Established. *Ndmpcopy: pacsnas1: Connect: Connection Established. *Ndmpcopy: alifiler: Log: DUMP: *Ndmpcopy: alifiler: Log: cannot open /vol/vol0/test. *Ndmpcopy: alifiler: Log: DUMP: *Ndmpcopy: alifiler: Log: DUMP IS ABORTED *Ndmpcopy: alifiler: Notify: dump failed *Ndmpcopy: pacsnas1: Notify: restore failed *Ndmpcopy: Transfer failed *Ndmpcopy: Fatal error. Aborting ... * *I can make a connection but for some reason I can't open the *file on the *source. Any suggestions before I upgrade would be appreciated. *Also, I ran *the same command with ndmpd debug set to 50 and have those results. If *someone out there would like to look at this maybe they can understand *what's happening better than I. E-mail me off the list and I *will send a *copy of the debug output if you're up to it. *Thanks again for everyone's help. *

1 0

OOPS!!
by Haber, David J. 18 Dec '02

18 Dec '02

I have changed the root password on my filers! Thanks for pointing out my egregious oversight!

1 0

Ndmpcopy
by Haber, David J. 18 Dec '02

18 Dec '02

Thanks to everyone for their responses. I am getting closer to a solution. The consensus from the NetApp engineers is that I need to upgrade the source filer (840) to at least version 6.2.1. Others have told me that I should be able to run ndmpcopy from a third machine and still be able to copy the files from the 840 to the NearStore even though the 840 is not at 6.2.1. This is where I'm at before the upgrade: NEARSTORE> ndmpcopy -sa root:dedsam9ap -da root:dedsam9ap -st text -dt text alifiler:/vol/vol0/test p acsnas1:/vol/vol0/test Ndmpcopy: Starting copy [ 21 ] ... Ndmpcopy: alifiler: Connect: Connection Established. Ndmpcopy: pacsnas1: Connect: Connection Established. Ndmpcopy: alifiler: Log: DUMP: Ndmpcopy: alifiler: Log: cannot open /vol/vol0/test. Ndmpcopy: alifiler: Log: DUMP: Ndmpcopy: alifiler: Log: DUMP IS ABORTED Ndmpcopy: alifiler: Notify: dump failed Ndmpcopy: pacsnas1: Notify: restore failed Ndmpcopy: Transfer failed Ndmpcopy: Fatal error. Aborting ... I can make a connection but for some reason I can't open the file on the source. Any suggestions before I upgrade would be appreciated. Also, I ran the same command with ndmpd debug set to 50 and have those results. If someone out there would like to look at this maybe they can understand what's happening better than I. E-mail me off the list and I will send a copy of the debug output if you're up to it. Thanks again for everyone's help.

1 0

RE: Ndmpcopy again
by Kennedy, Jeffrey 18 Dec '02

18 Dec '02

I believe the problem is that ndmpcopy was only introduced as part of OT in 6.2; previously it required a platform specific binary. The original poster's 840 filer is running 6.1.1. ~JK > -----Original Message----- > From: Collin Park [mailto:collin@netapp.com] > Sent: Wednesday, December 18, 2002 7:47 AM > To: David J. Haber > Cc: 'toasters(a)mathworks.com' > Subject: Re: Ndmpcopy again > > >This is what I got: > >Ndmpcopy: starting copy [13] > >Ndmpcopy: Trouble connecting to destination NearStore > >Ndmpcopy: fatal error. Aborting.... > > Does your version of software support "ndmpd debug 50" ? If so, > issuing that command on both sides should tell you what IP address[es] > are being used/recommended, and at what point the connection isn't > connecting. > > === > Collin Park Not a statement of my employer. > 408-822-6878 3.1.039

2 1

RE: Ndmpcopy again
by Noll, Kevin 18 Dec '02

18 Dec '02

Dont know if this helps, but here is my how to on NDMP Task NDMP Copy Commands: options ndmpd Outputs the options for ndmp. Ndmpd status Status of NDMP Ndmpd on/off Switches ndmp on / off Options ndmpd.enable on Enables NDMP Options ndmpd.enable off Disables NDMP Options ndmpd.preferred_interface Specifies the interface through which ndmp would send data Steps to enable NDMP Options ndmpd.enable on Options ndmpd.preferred_interface <interface name> Commands to copy data through NDMP Volumes, Qtrees, Directories can be copied through NDMP using the below syntax ndmpcopy -sa username:password -da username:password <source> <destination> where sa and da are source/destination filer authentication <source> and <destination> are of the form [<filer>:]<path> Limitations: Individual file(s) cannot be copied through NDMP Not preferred for transfer for small data Kevin Noll tel +44 7801765042 Cap Gemini Ernst & Young -----Original Message----- From: Kennedy, Jeffrey [mailto:jkennedy@qualcomm.com] Sent: 18 December 2002 16:01 To: Collin Park; David J. Haber Cc: toasters(a)mathworks.com Subject: RE: Ndmpcopy again I believe the problem is that ndmpcopy was only introduced as part of OT in 6.2; previously it required a platform specific binary. The original poster's 840 filer is running 6.1.1. ~JK > -----Original Message----- > From: Collin Park [mailto:collin@netapp.com] > Sent: Wednesday, December 18, 2002 7:47 AM > To: David J. Haber > Cc: 'toasters(a)mathworks.com' > Subject: Re: Ndmpcopy again > > >This is what I got: > >Ndmpcopy: starting copy [13] > >Ndmpcopy: Trouble connecting to destination NearStore > >Ndmpcopy: fatal error. Aborting.... > > Does your version of software support "ndmpd debug 50" ? If so, > issuing that command on both sides should tell you what IP address[es] > are being used/recommended, and at what point the connection isn't > connecting. > > === > Collin Park Not a statement of my employer. > 408-822-6878 3.1.039 ******************************************************************************************** " This message contains information that may be privileged or confidential and is the property of the Cap Gemini Ernst & Young Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message ". ********************************************************************************************

1 0

Ndmpcopy again
by Haber, David J. 18 Dec '02

18 Dec '02

I used this command from the Nearstore ndmpcopy -sa root:passwd -da root:passwd 840:/vol/vol0/test NearStore:/vol/vol0/ (I used the real names for the 840 and the Nearstore) This is what I got: Ndmpcopy: starting copy [13] Ndmpcopy: Trouble connecting to destination NearStore Ndmpcopy: fatal error. Aborting.... Ndmpcopy: Done

3 2

test1
by Roy Ruigrok 18 Dec '02

18 Dec '02

this is a test, please ignore __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

1 0

Ndmpcopy
by Haber, David J. 18 Dec '02

18 Dec '02

I have to move 60G of data from a folder on an 840 filer to a folder on a NearStore. I want to use ndmpcopy however, I can't seem to get it to work. The 840 is running 6.1.1R1 and the NearStore is running 6.2.1. Does the 840 need to be upgraded to 6.2.1? Also, there is no ndmpcopy on the 840. Is this because of the downlevel version of software? NetApp tech support has not been helpful at all. I was on the phone with them about this issue for a few hours today and they couldn't answer these questions. Also, I think I got some bad advice on how to set up the /etc/hosts file and the /etc/hosts.equiv file. Any help would be appreciated.

1 0

RE: CIFS Files are Moving
by Steve Evans 17 Dec '02

17 Dec '02

I just talked with the help desk guy earlier this morning and I just found out that it has happening in different departments, and they are headed to different locations. Also nobody has an account that has access to all the different places this has been happening. Also nobody here has Windows XP. Steve Evans SDSU Foundation (619) 594-0653 -----Original Message----- From: Joseph Bishop [mailto:Joseph.Bishop@jpl.nasa.gov] Sent: Monday, December 16, 2002 9:39 PM To: Steve Evans; toasters(a)mathworks.com Subject: Re: CIFS Files are Moving Steve, This might be reaching a little, but who knows. I have noticed on my WinXp machine that if I hover with the mouse over a directory name in Windows Explorer, that it will auto select the directory. It is possible that the person that has an access to all users directories could be auto-selecting the directory....then thinking they are clicking on another directory on the same level, they actually dragged one directory under another. I don't like this handy feature in Windows...it can be a problem if you are not used to it. I am used to having to click on the directory to highlight, then hold down the mouse to drag, then release on the directory you want to drop/move the folder to. It is not to hard to accidentally do this if the auto select feature is on. I have only noticed this problem on my personal machine...and I don't know how to turn off the auto select. Maybe that will help...maybe it is not related. Sounds like someone is accidentally moving folders though. Someone with the right level of access. Cheers! Joe At 07:56 PM 12/16/02 -0800, Steve Evans wrote: >We have one department that is having a problem with their share on our >Filer. > >About 4 times in the last month a users directory has been randomly >placed in a certain users folder. So two folders that are usually on >the same level (\\filer\dept\user-a and \\filer\dept\user-b) all of a >sudden user-a will be under user-b. The other interesting thing is >that when you go to \\filer\dept\user-a\~snapshot\ all of the folders >are listed as files. (named hourly.0, weekly.0, etc) But if you go >through \\filer\dept\~snapshot\hourly.0\user-a everything is fine. We >are almost sure that this isn't a user doing something stupid (but we >can't be sure.) > >Steve Evans >SDSU Foundation >(619) 594-0653

1 0

Re: CIFS Files are Moving
by Joseph Bishop 17 Dec '02

17 Dec '02

Steve, This might be reaching a little, but who knows. I have noticed on my WinXp machine that if I hover with the mouse over a directory name in Windows Explorer, that it will auto select the directory. It is possible that the person that has an access to all users directories could be auto-selecting the directory....then thinking they are clicking on another directory on the same level, they actually dragged one directory under another. I don't like this handy feature in Windows...it can be a problem if you are not used to it. I am used to having to click on the directory to highlight, then hold down the mouse to drag, then release on the directory you want to drop/move the folder to. It is not to hard to accidentally do this if the auto select feature is on. I have only noticed this problem on my personal machine...and I don't know how to turn off the auto select. Maybe that will help...maybe it is not related. Sounds like someone is accidentally moving folders though. Someone with the right level of access. Cheers! Joe At 07:56 PM 12/16/02 -0800, Steve Evans wrote: >We have one department that is having a problem with their share on our >Filer. > >About 4 times in the last month a users directory has been randomly >placed in a certain users folder. So two folders that are usually on >the same level (\\filer\dept\user-a and \\filer\dept\user-b) all of a >sudden user-a will be under user-b. The other interesting thing is that >when you go to \\filer\dept\user-a\~snapshot\ all of the folders are >listed as files. (named hourly.0, weekly.0, etc) But if you go through >\\filer\dept\~snapshot\hourly.0\user-a everything is fine. We are >almost sure that this isn't a user doing something stupid (but we can't >be sure.) > >Steve Evans >SDSU Foundation >(619) 594-0653

1 0

CIFS Files are Moving
by Steve Evans 17 Dec '02

17 Dec '02

We have one department that is having a problem with their share on our Filer. About 4 times in the last month a users directory has been randomly placed in a certain users folder. So two folders that are usually on the same level (\\filer\dept\user-a and \\filer\dept\user-b) all of a sudden user-a will be under user-b. The other interesting thing is that when you go to \\filer\dept\user-a\~snapshot\ all of the folders are listed as files. (named hourly.0, weekly.0, etc) But if you go through \\filer\dept\~snapshot\hourly.0\user-a everything is fine. We are almost sure that this isn't a user doing something stupid (but we can't be sure.) Steve Evans SDSU Foundation (619) 594-0653

1 0

Netapp or Oracle problems every sunday....
by Art Hebert 17 Dec '02

17 Dec '02

We are having a problem with out netapp or oracle config every sunday. Is there a way to find out what if anything is running on a netapp at a scheduled time? eg. disk scrubs on sundays Thanks art hebert

2 1

mounting foreign volumes as readonly
by steve＠hq.newdream.net 16 Dec '02

16 Dec '02

Is it possible to mount a foreign volume as readonly on a different filer? I want to mount a corrupted volume as readonly on another working filer. In fact, I can only get this corrupted volume to work when its in that readonly mode and using inodes from an old snapshot. If the corrupted volume boots normal it causes the filer to panick. So I guess what I'm wondering is if there a way to boot up one volume normally and another volume as readonly (using old inodes)? Steve

1 0

case changing CIFS filenames
by p dar 16 Dec '02

16 Dec '02

I'm considering moving some users from an unsupported Samba server, to our F760. However it's not clear that the filer can accurately replace all the Samba features. Some of my users have become accustomed to relying on the fact that our Samba server allows me to either force all CIFS filenames to lower case, or preserve the case, both of which aren't part of the CIFS protocol. Isthere a way for me to make my filer preserve the case in which files are created via CIFS? Thus allowing users to create and access 2 different files named Test.txt and test.txt. These are the Samba options I'm using and wondering if there's an equivalent OnTap configuration? ; Preserve filenames and their case case sensitive = yes preserve case = yes short preserver case = yes Alternatively, can I make my filer force CIFS created files to lower-case? [Default Samba behaviour]? Are there any other gotcha's or features that the Samba server offers that may not be available on my filer? I want to be sure I can keep these users happy before I move them. Cheers, Mark bg_canuckNO-SPAM-HERE(a)yahoo.com <Remove the Capital letters and Hyphens if you need to send me email directly> CIRT - Center for Internet Resources and Technology --------------------------------- Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now

2 1

CIFS and LDAP support
by Frank Bonnet 16 Dec '02

16 Dec '02

Hi Actually we use samba servers to let our windows users access to their data, samba servers act also as wins servers and PDC servers. Samba use NFS to access the filer. I know this is FAR to be the best practice, but Ontap still does not support native LDAP authentication ( Openldap ). I have requested several times in this list WHEN Ontap will support LDAP auth but never received a positive answer so I repost my request ... I actually runs 6.3.1 Ontap version. PS : I do not wants to use M$ Active Directory as our LDAP server runs under UNIX of course. Thanks for any news. -- Frank Bonnet

1 0

Experiences with filers running NFS over long distance
by Garrett Burke 14 Dec '02

14 Dec '02

Hi All, I'm wondering if anyone out there has experience with running NFS with filers over long distances - of the order of 10s of KM. I'm particularly interested to find out experiences when bandwidth was and wasn't an issue. For example, running at 34Mbs/155Mbs against running over GigE. Was tweaking the NFS request size required? Was TCP used and if so was it because pkt loss was impacting NFS performance? Did latency become an issue? Thanks, GB ---- Garrett Burke, Service Implementation & Support Manager, Eircomnet. http://www.eircom.net

3 2

未承諾広告※必見！裏情報！
by jmail5555jp 13 Dec '02

13 Dec '02

<MÒ> dq[LÐ ¡ãALð²ó]³êÈ¢ûÍ±±Ö info_ssy(a)yahoo.co.jp Kž{¶É ÈœÌ[AhXÌÝðš«º³¢ =============================================================== ÐÌåÈLÍð©µÜ¹ñ©I zMÆ±©çz[y[W»ìÜÅiÀÉÄšó¯vµÜ·B ºLe`wÉÄš\µÝº³¢B =============================================================== §104-0061 sæâÀ8-19-3 æ2ECOr@3F [}KWs TEL@03-3544-6222 FAX@03-3544-6218@@ =============================================================== âè€iÎ©èWßÜµœBÁ³êé°êª èÜ·ÌÅ š\ÝÍšßÉI ================================================================= \\\\\\\\\\\\\\\\\\\\\\\\ rfIÌEÁê_b`CtErlNu @@ `ujDåWEðÛErdwthEA_gObYÈÇ @A_gÖAÌîñÚ@ @@š\ÝE²¶E€iÚ×Í@ @@@@@ºLtqkðNbNµÄ²º³¢B «@@@@«@@@@«@ @@@http://www.xx-koukoku.com/ \\\\\\\\\\\\\\\\\\\\\\\\ @@J^ObYEÉéîñEhÆObYEà×¯îñÈÇ@ @@@@@@@@@@»ÌŒîñÚ@ @@š\ÝE²¶E€iÚ×Í@ @@@@@ºLtqkðNbNµÄ²º³¢B «@@@@«@@@@«@ http://www.ww-koukoku.com/ \\\\\\\\\\\\\\\\\\\\\\\\

1 0

cloning of filer config
by Thorsten.Lemke＠GECITS-EU.COM 13 Dec '02

13 Dec '02

Hi, does anybody have some experiences with the cloning of a filers config? What exactly are the configurations that are cloned? We have two F840 and one of them is used to snapmirror every volume of his partner except the root volume. Can we use the clone command to sync the configuration files on both filers? If we need to break up the mirrors in case of emergency, the filer should be able to act as an exact clone of the source filer with all nfs and cifs exports, options, permissions etc. Interfaces of course are already configured the same. Regards Thorsten Lemke

1 0

Re: file folding impact
by Timothy Cook 13 Dec '02

13 Dec '02

As I understand it, with windoze systems, when you modify a file, you completely rewrite it as a new file. Those blocks that belonged to the "old" file are maintained in the snapshot inode listing. This is not an issue with Unix (As I understand it). So with file folding, the contents of the active file system are compared against the most recent snapshot (only) to determine if blocks are duplicates and can be freed. Also, small files (less than 64K, which really are written to the inode) NT streams and directories are not folded. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ This electronic mail transmission contains information from Joy Mining Machinery which is confidential, and is intended only for the use of the proper addressee. If you are not the intended recipient, please notify us immediately at the return address on this transmission, or by telephone at (724) 779-4500, and delete this message and any attachments from your system. Unauthorized use, copying, disclosing, distributing, or taking any action in reliance on the contents of this transmission is strictly prohibited and may be unlawful. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

3 2

netgroup and user field
by Frank Bonnet 13 Dec '02

13 Dec '02

Hi I noticed that the "user" field is ignored by Ontap in the netgroup file according I use "files" only in the nsswitch.conf file. (machine,user,domain) ^^^^ Is there some plan at Netapp to support this field in the future to let only some users access to the filer from certain machines ? Thanks -- Frank Bonnet

1 0

Re: file folding impact --Correction!!!
by Timothy Cook 13 Dec '02

13 Dec '02

I misspoke. small files are files < 64 bytes, (NOT kb!) Oops! As I understand it, with windoze systems, when you modify a file, you completely rewrite it as a new file. Those blocks that belonged to the "old" file are maintained in the snapshot inode listing. This is not an issue with Unix (As I understand it). So with file folding, the contents of the active file system are compared against the most recent snapshot (only) to determine if blocks are duplicates and can be freed. Also, small files (less than 64K, which really are written to the inode) NT streams and directories are not folded. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ This electronic mail transmission contains information from Joy Mining Machinery which is confidential, and is intended only for the use of the proper addressee. If you are not the intended recipient, please notify us immediately at the return address on this transmission, or by telephone at (724) 779-4500, and delete this message and any attachments from your system. Unauthorized use, copying, disclosing, distributing, or taking any action in reliance on the contents of this transmission is strictly prohibited and may be unlawful. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0

RE: file folding impact
by Chuck Tomasi 12 Dec '02

12 Dec '02

My question about file folding (after reading about it here and on NOW)... Why would a block end up in a snapshot if it is the same as the previous snapshot? Isn't the point of snapshots only to save the 'old version' of blocks that have been changed? OK, second question, does it only check the most recent snapshot or does it drill through all existing snaps? Now that I think more on the subject, third question... what happens to the snapshot "n" that references blocks in snap "n+1" when "n+1" rolls off the end and expires? --Chuck -----Original Message----- From: Mitko Blazeski [mailto:Mitko.Blazeski@proact.se] Sent: Thursday, December 12, 2002 3:10 AM To: 'Kumar, Rahul'; 'toasters(a)mathworks.com' Subject: RE: file folding impact File folding describes the process of checking the data in the most recent snapshot, and, if it is identical to the snapshot currently being created, just referencing the previous snapshot instead of taking up disk space writing the same data in the new snapshot. Disk space is saved by sharing unchanged file blocks between the active version of the file and the version of the file in the latest snapshot, if any. /Mitko -----Original Message----- From: Kumar, Rahul [mailto:rahul.kumar@eds.com] Sent: den 12 december 2002 05:19 To: stefan.holzwarth(a)adac.de; mats.oberg(a)tietoenator.com; toasters(a)mathworks.com Subject: RE: file folding impact BTW what does this option do Rahul -----Original Message----- From: stefan.holzwarth(a)adac.de [mailto:stefan.holzwarth@adac.de] Sent: Wednesday, December 11, 2002 7:08 PM To: mats.oberg(a)tietoenator.com; toasters(a)mathworks.com Subject: AW: file folding impact Since 2 months we use filefolding on all volumes. (F760,2TB,3000 User, only cifs) No remarkable impact in cpu usage. No trouble. Snapshotsize decreasing a little. Regards Stefan Holzwarth -----Ursprüngliche Nachricht----- Von: Öberg Mats [mailto:mats.oberg@tietoenator.com] Gesendet: Mittwoch, 11. Dezember 2002 14:20 An: toasters(a)mathworks.com Betreff: file folding impact Hi, I'm thinking about enabling file folding on one of our filers. Have anybody tried this? If so, what kind of performance decrease/size gain can be expected from enabling it? ---- Mats

4 4

RE: NFS Lock Performance
by Edward Hibbert 12 Dec '02

12 Dec '02

No, I have multiple client systems, so I do need the locks to go to the NetApp. Edward. -----Original Message----- From: Duncan Greenwood [mailto:duncan.greenwood@btinternet.com] Sent: 12 December 2002 22:56 To: Edward Hibbert; toasters(a)mathworks.com Subject: Re: NFS Lock Performance You don't say whether you have multiple client systems mounting the share, or just a single client with multiple processes contending for locks. In the latter situation and depending on your client o/s, you may be able to dispense with the lock manager altogether and use local kernel locks instead, using the "llock" option to the NFS mount command. rgds, D # ----- Original Message ----- From: "Edward Hibbert" <EH(a)dataconnection.com> To: <toasters(a)mathworks.com> Sent: Thursday, December 12, 2002 10:13 PM Subject: NFS Lock Performance > I'm hitting a filer very hard with NFS lock/unlock calls. Although the > ops/second is still comfortably below what I've seen it do before (with more > read/write ops), I'm seeing it occasionally take many seconds to return from > a non-blocking lock attempt, or from an unlock call. > > I theorise based purely on ignorance that I might be soaking the lock > manager. Is there any tuning I can do (e.g. add more threads)? > > Regards, > > Edward Hibbert. --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.419 / Virus Database: 235 - Release Date: 13/11/2002

1 0

NFS Lock Performance
by Edward Hibbert 12 Dec '02

12 Dec '02

I'm hitting a filer very hard with NFS lock/unlock calls. Although the ops/second is still comfortably below what I've seen it do before (with more read/write ops), I'm seeing it occasionally take many seconds to return from a non-blocking lock attempt, or from an unlock call. I theorise based purely on ignorance that I might be soaking the lock manager. Is there any tuning I can do (e.g. add more threads)? Regards, Edward Hibbert.

2 1

Jump to page:

2026

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