Re: [Lustre-discuss] Lustre 1.8.1 distribution missing kernel source?

2009-08-17 Thread Richard Smith
Brian J. Murrell wrote: Because you didn't find a kernel-source, I am guessing you are using the RHEL5 packages. RH doesn't package a kernel-source package, only Suse does. RH requires that you get the source from the src.rpm. There's tonnes of info on the web on how to build a kernel from

Re: [Lustre-discuss] 1.8.1 kernel-lustre rpm installation failed

2009-08-17 Thread CHU, STEPHEN H, ATTSI
Hi Felix, Thanks for the pointer. I will upgrade to the matching kernels and go from there. Steve -Original Message- From: Felix Frank [mailto:felix.fr...@desy.de] Sent: Sunday, August 16, 2009 8:56 AM To: CHU, STEPHEN H, ATTSI Cc: lustre-discuss@lists.lustre.org Subject: Re:

Re: [Lustre-discuss] Lustre 1.8.1 distribution missing kernel source?

2009-08-17 Thread Brian J. Murrell
On Mon, 2009-08-17 at 17:31 +1000, Richard Smith wrote: Yes, I was using RHEL5 x86_64 packages. FWIW I did go and compile the kernel after installing the src.rpm, via rpmbuild -bb kernel-2.6.spec. Yes, IIRC that is one of the possible paths that RH describe. If you just wanted patched source

Re: [Lustre-discuss] Why are there many threads named ll_imp_invalin client?

2009-08-17 Thread huangql
Hi, Alexey The version of lustre we are using is V1.6.6. I hope you can give me more suggestions. Thanks, Sarea 2009-08-17 huangql 发件人: Alexey Lyashkov 发送时间: 2009-08-17 22:31:58 收件人: huangql 抄送: lustre-discuss 主题: Re: [Lustre-discuss] Why are there many threads named

[Lustre-discuss] [Fwd: [ofa-general] IPoIB Transmit Timeouts]

2009-08-17 Thread Charles A. Taylor
FWIW, I posted this to ofa-general a little earlier. Anyone else seeing this?Suggestions?I think this is an OFED 1.4.1 problem but they may point the finger at you guys. :) We've tried limiting OST threads to no avail. It doesn't really seem to require a heavy load to trigger it -

Re: [Lustre-discuss] MDS refuses connections (no visible reason)

2009-08-17 Thread Patricia Santos Marco
The last day our MDS refusing conections too. The logs are the same, and we should reboot the MDS server . What's is the reason for this? 2009/3/5 Thomas Roth t.r...@gsi.de Hi all, after running for days without any problems, our MDS is refusing cooperation for two hours now. The log files

Re: [Lustre-discuss] [Fwd: [ofa-general] IPoIB Transmit Timeouts]

2009-08-17 Thread Nirmal Seenu
I was getting these same errors when I was running the following kernel: kernel-lustre-smp-2.6.18-92.1.17.el5_lustre.1.6.7.1.x86_64 These errors went away when I started using 2.6.22.19 with lustre patches + OFED-1.4.2 (http://www.openfabrics.org/downloads/OFED/ofed-1.4.2/OFED-1.4.2.tgz) on

Re: [Lustre-discuss] Lustre 1.8.1 distribution missing kernel source?

2009-08-17 Thread Andreas Dilger
On Aug 17, 2009 09:44 -0400, Brian J. Murrell wrote: On Mon, 2009-08-17 at 17:31 +1000, Richard Smith wrote: Yes, I was using RHEL5 x86_64 packages. FWIW I did go and compile the kernel after installing the src.rpm, via rpmbuild -bb kernel-2.6.spec. Yes, IIRC that is one of the possible

Re: [Lustre-discuss] Lustre 1.8.1 distribution missing kernel source?

2009-08-17 Thread Brian J. Murrell
On Mon, 2009-08-17 at 12:18 -0600, Andreas Dilger wrote: Yes, IIRC that is one of the possible paths that RH describe. If you just wanted patched source which you could then further patch/tweak before you executed the build, you could just use rpmbuild -bp, which I believe is one of

Re: [Lustre-discuss] MDS refuses connections (no visible reason)

2009-08-17 Thread Oleg Drokin
Hello! On Aug 17, 2009, at 2:14 PM, Patricia Santos Marco wrote: The last day our MDS refusing conections too. The logs are the same, and we should reboot the MDS server . What's is the reason for this? That means some requests from this client are still being processed and server has a

[Lustre-discuss] Infiband to TCP router

2009-08-17 Thread Aaron Lauer
We have Lustre test lab configured using 3 OSS and 1 MGS/MDS connected over Infiniband and ethernet. How do you configure a LNET router to allow non IB servers to access Lustre over TCP? Aaron Lauer IT Engineer Digitalsmiths www.digitalsmiths.com http://www.digitalsmiths.com (O)

[Lustre-discuss] routing between infiniband and TCP networks

2009-08-17 Thread Aaron Lauer
I have a small Lustre setup consisting of 3xOSS and 1xMDS/MGS connected via Infiniband. How do you configure a LNET router to route between the IB and TCP networks so a non-IB client can connect and use Lustre? Our IB network is 192.168.200. and our TCP network is 192.168.10. Thanks,

Re: [Lustre-discuss] [Fwd: [ofa-general] IPoIB Transmit Timeouts]

2009-08-17 Thread Isaac Huang
On Mon, Aug 17, 2009 at 12:23:35PM -0400, Charles A. Taylor wrote: FWIW, I posted this to ofa-general a little earlier. Anyone else seeing this?Suggestions?I think this is an OFED 1.4.1 problem but they may point the finger at you guys. :) We've tried limiting OST threads to no

Re: [Lustre-discuss] Infiband to TCP router

2009-08-17 Thread Brian J. Murrell
On Mon, 2009-08-17 at 17:21 -0400, Aaron Lauer wrote: We have Lustre test lab configured using 3 OSS and 1 MGS/MDS connected over Infiniband and ethernet. How do you configure a LNET router to allow non IB servers to access Lustre over TCP? There is a whole section of the manual

Re: [Lustre-discuss] Infiband to TCP router

2009-08-17 Thread Andreas Dilger
On Aug 17, 2009 17:21 -0400, Aaron Lauer wrote: We have Lustre test lab configured using 3 OSS and 1 MGS/MDS connected over Infiniband and ethernet. How do you configure a LNET router to allow non IB servers to access Lustre over TCP? Have you tried reading the manual? manual.lustre.org

Re: [Lustre-discuss] [Fwd: [ofa-general] IPoIB Transmit Timeouts]

2009-08-17 Thread Craig Prescott
Isaac Huang wrote: On Mon, Aug 17, 2009 at 12:23:35PM -0400, Charles A. Taylor wrote: FWIW, I posted this to ofa-general a little earlier. Anyone else seeing this?Suggestions?I think this is an OFED 1.4.1 problem but they may point the finger at you guys. :) We've tried limiting

Re: [Lustre-discuss] Lustre 1.8.1 distribution missing kernel source?

2009-08-17 Thread Richard Smith
Brian J. Murrell wrote: Nope. rpmbuild -bb should have built the lustre kernel RPMs. What did get created? Here's the list of artifacts that was created in /usr/src/redhat/RPMS/x86_64: kernel-2.6.18-128.1.14.el5.x86_64.rpm kernel-debug-2.6.18-128.1.14.el5.x86_64.rpm

Re: [Lustre-discuss] Why are there many threads named ll_imp_invalin client?

2009-08-17 Thread huangql
Hi, all Sorry, I should show the lustre version in details. The version for our system as follows: MDS:V1.6.5 OSS:V1.6.6 Clients: V1.6.6 Thank you for your help. Best wishes, Sarea 2009-08-18 huangql 发件人: Alexey Lyashkov 发送时间: 2009-08-17 22:31:58 收件人: huangql 抄送: