From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Duv Dek" Subject: NFS response time Date: Tue, 25 Jan 2005 23:40:12 +0200 Message-ID: <003a01c50326$73182550$7889b38b@alpin> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0037_01C50337.3675EDB0" Return-path: Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.11] helo=sc8-sf-mx1.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CtYQL-0000B0-LE for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 13:40:33 -0800 Received: from dalyan.ee.bilkent.edu.tr ([139.179.12.25]) by sc8-sf-mx1.sourceforge.net with esmtp (Exim 4.41) id 1CtYQK-00012g-Ef for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 13:40:33 -0800 Received: from alpin (d79012c.dorm.bilkent.edu.tr [139.179.137.120]) by dalyan.ee.bilkent.edu.tr (8.12.9/8.12.9) with SMTP id j0PLeFtf020511 for ; Tue, 25 Jan 2005 23:40:15 +0200 (EET) To: Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: This is a multi-part message in MIME format. ------=_NextPart_000_0037_01C50337.3675EDB0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi We have a problem with our NFS server on Redhat 9. In case a program on = the server uses high CPU, the NFS clients almost lock and nfs stops = giving response until that program lowers its CPU usage. We tried to = increase the priority of nfsd deamons with renice, and also tried to = increase the number of nfsd threads, decreased the priority of the = program using high CPU with renice, but none of them worked. How can we = solve this problem so that nfs does not lock? ------=_NextPart_000_0037_01C50337.3675EDB0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
We have a problem with our NFS server = on Redhat 9.=20 In case a program on the server uses high CPU, the=20 NFS clients almost lock and nfs stops giving response until that = program=20 lowers its CPU usage.  We tried to increase the priority of = nfsd=20 deamons with renice, and also tried to increase the number of nfsd = threads,=20 decreased the priority of the program using high CPU with = renice, but none=20 of them worked. How can we solve this problem so that nfs does not=20 lock?
 
------=_NextPart_000_0037_01C50337.3675EDB0-- ------------------------------------------------------- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag-&-drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Duv Dek" Subject: Re: NFS response time Date: Wed, 26 Jan 2005 00:00:11 +0200 Message-ID: <004101c50329$3e016b30$7889b38b@alpin> References: <003a01c50326$73182550$7889b38b@alpin> <41F6BD96.4040409@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=response Cc: Return-path: Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.11] helo=sc8-sf-mx1.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CtYjZ-0001Bz-QA for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 14:00:25 -0800 Received: from dalyan.ee.bilkent.edu.tr ([139.179.12.25]) by sc8-sf-mx1.sourceforge.net with esmtp (Exim 4.41) id 1CtYjW-00039a-HB for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 14:00:25 -0800 To: "Neil Horman" Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: Yes, it is program that we wrote. It is doing some mathematical calculations, so it utilizes full CPU. Other linux programs are not getting very slow when this program runs, but client computers sometimes wait half a minute in order to reach a file on nfs server. Logging in to the client computers take around 10 minutes. Normally it takes no more than 20 seconds. I mean nfs is almost locking, but no other linux program is badly affected. At least we should be able to give better priority nfs via renice but we couldn't figure it out. Doing renice -20 to all root programs didn't make nfs faster. ----- Original Message ----- From: "Neil Horman" To: "Duv Dek" Cc: Sent: Tuesday, January 25, 2005 11:43 PM Subject: Re: [NFS] NFS response time > Duv Dek wrote: >> Hi >> We have a problem with our NFS server on Redhat 9. In case a program on >> the server uses high CPU, the NFS clients almost lock and nfs stops >> giving response until that program lowers its CPU usage. We tried to >> increase the priority of nfsd deamons with renice, and also tried to >> increase the number of nfsd threads, decreased the priority of the >> program using high CPU with renice, but none of them worked. How can we >> solve this problem so that nfs does not lock? >> > Have your tried to ascertain which process on your system is responsible > (or most responsible) for the high CPU utilization? > > Neil > > -- > /*************************************************** > *Neil Horman > *Software Engineer > *Red Hat, Inc. > *nhorman@redhat.com > *gpg keyid: 1024D / 0x92A74FA1 > *http://pgp.mit.edu > ***************************************************/ > ------------------------------------------------------- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag-&-drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Duv Dek" Subject: NFS response time Date: Tue, 25 Jan 2005 19:19:12 +0200 Message-ID: <007a01c50301$fd199510$7889b38b@alpin> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0077_01C50312.C05619D0" Received: from sc8-sf-mx2-b.sourceforge.net ([10.3.1.12] helo=sc8-sf-mx2.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CtULl-0003uK-Ua for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 09:19:33 -0800 Received: from dalyan.ee.bilkent.edu.tr ([139.179.12.25]) by sc8-sf-mx2.sourceforge.net with esmtp (Exim 4.41) id 1CtULj-00039R-JV for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 09:19:33 -0800 Received: from alpin (d79012c.dorm.bilkent.edu.tr [139.179.137.120]) by dalyan.ee.bilkent.edu.tr (8.12.9/8.12.9) with SMTP id j0PHJFtf017691 for ; Tue, 25 Jan 2005 19:19:15 +0200 (EET) To: Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: This is a multi-part message in MIME format. ------=_NextPart_000_0077_01C50312.C05619D0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi We have a problem with our NFS server on Redhat 9. In case a program on = the server uses high CPU, the NFS clients almost lock and nfs stops = giving response until that program lowers its CPU usage. We tried to = increase the priority of nfsd deamons with renice, and also tried to = increase the number of nfsd threads, decreased the priority of the = program using high CPU with renice, but none of them worked. How can we = decrease this problem so that nfs does not lock? ------=_NextPart_000_0077_01C50312.C05619D0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
We have a problem with our NFS server = on Redhat 9.=20 In case a program on the server uses high CPU, the=20 NFS clients almost lock and nfs stops giving response until that = program=20 lowers its CPU usage.  We tried to increase the priority of = nfsd=20 deamons with renice, and also tried to increase the number of nfsd = threads,=20 decreased the priority of the program using high CPU with = renice, but none=20 of them worked. How can we decrease this problem so that nfs does = not=20 lock?
------=_NextPart_000_0077_01C50312.C05619D0-- ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Horman Subject: Re: NFS response time Date: Tue, 25 Jan 2005 16:43:50 -0500 Message-ID: <41F6BD96.4040409@redhat.com> References: <003a01c50326$73182550$7889b38b@alpin> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Cc: nfs@lists.sourceforge.net Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.11] helo=sc8-sf-mx1.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CtYTf-0000RQ-RL for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 13:43:59 -0800 Received: from mx1.redhat.com ([66.187.233.31]) by sc8-sf-mx1.sourceforge.net with esmtp (TLSv1:AES256-SHA:256) (Exim 4.41) id 1CtYTe-0001UJ-CZ for nfs@lists.sourceforge.net; Tue, 25 Jan 2005 13:43:59 -0800 To: Duv Dek In-Reply-To: <003a01c50326$73182550$7889b38b@alpin> Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: Duv Dek wrote: > Hi > > We have a problem with our NFS server on Redhat 9. In case a program on > the server uses high CPU, the NFS clients almost lock and nfs stops > giving response until that program lowers its CPU usage. We tried to > increase the priority of nfsd deamons with renice, and also tried to > increase the number of nfsd threads, decreased the priority of the > program using high CPU with renice, but none of them worked. How can > we solve this problem so that nfs does not lock? > Have your tried to ascertain which process on your system is responsible (or most responsible) for the high CPU utilization? Neil -- /*************************************************** *Neil Horman *Software Engineer *Red Hat, Inc. *nhorman@redhat.com *gpg keyid: 1024D / 0x92A74FA1 *http://pgp.mit.edu ***************************************************/ ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Horman Subject: Re: NFS response time Date: Wed, 26 Jan 2005 07:46:54 -0500 Message-ID: <41F7913E.9030404@redhat.com> References: <003a01c50326$73182550$7889b38b@alpin> <41F6BD96.4040409@redhat.com> <004101c50329$3e016b30$7889b38b@alpin> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Cc: nfs@lists.sourceforge.net Received: from sc8-sf-mx2-b.sourceforge.net ([10.3.1.12] helo=sc8-sf-mx2.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CtmZh-0003lw-QQ for nfs@lists.sourceforge.net; Wed, 26 Jan 2005 04:47:09 -0800 Received: from mx1.redhat.com ([66.187.233.31]) by sc8-sf-mx2.sourceforge.net with esmtp (TLSv1:AES256-SHA:256) (Exim 4.41) id 1CtmZg-00033v-5a for nfs@lists.sourceforge.net; Wed, 26 Jan 2005 04:47:09 -0800 To: Duv Dek In-Reply-To: <004101c50329$3e016b30$7889b38b@alpin> Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: Duv Dek wrote: > Yes, it is program that we wrote. It is doing some mathematical > calculations, so it utilizes full CPU. Other linux programs are not > getting very slow when this program runs, but client computers sometimes > wait half a minute in order to reach a file on nfs server. Logging in to > the client computers take around 10 minutes. Normally it takes no more > than 20 seconds. I mean nfs is almost locking, but no other linux > program is badly affected. At least we should be able to give better > priority nfs via renice but we couldn't figure it out. Doing renice -20 > to all root programs didn't make nfs faster. > I assume that you included your program in the renice? If it still didn't help, my immediate thought would be to enable nfsd_debug, try to access the NFS share from a client, and check the message log. That would at least tell you if the NFS server is receiving the NFS transactions, and if so, where they might be taking the most time in processing. That may give you a clue as to where the delay is comming from, be it a scheduling delay from high CPU load, or contention on another resource that is in short supply on your system. Neil > > > ----- Original Message ----- From: "Neil Horman" > To: "Duv Dek" > Cc: > Sent: Tuesday, January 25, 2005 11:43 PM > Subject: Re: [NFS] NFS response time > > >> Duv Dek wrote: >> >>> Hi >>> We have a problem with our NFS server on Redhat 9. In case a program >>> on the server uses high CPU, the NFS clients almost lock and nfs >>> stops giving response until that program lowers its CPU usage. We >>> tried to increase the priority of nfsd deamons with renice, and also >>> tried to increase the number of nfsd threads, decreased the priority >>> of the program using high CPU with renice, but none of them worked. >>> How can we solve this problem so that nfs does not lock? >>> >> Have your tried to ascertain which process on your system is >> responsible (or most responsible) for the high CPU utilization? >> >> Neil >> >> -- >> /*************************************************** >> *Neil Horman >> *Software Engineer >> *Red Hat, Inc. >> *nhorman@redhat.com >> *gpg keyid: 1024D / 0x92A74FA1 >> *http://pgp.mit.edu >> ***************************************************/ >> > -- /*************************************************** *Neil Horman *Software Engineer *Red Hat, Inc. *nhorman@redhat.com *gpg keyid: 1024D / 0x92A74FA1 *http://pgp.mit.edu ***************************************************/ ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs