similar to: options to use jumbo frames

Hi Brock, I'm trying set full duplex on a server that is being pxeboot/provisioned via autoyast. We want to ensure the duplex is full DURING the autoyast provisioning itself. I've played the syslinux configs to specify the bcm5600 driver, but was having trouble with setting the network options. I've tryed appending them as well, but nothing on the internet seems to indicate that

I use gentoo on xen, in the gentoo docs (gentoo compiles everything almost) they have options like ''-march=pentium3'' should i be using the gcc option for the cpu im using (pentium3), I understand that domU''s and even dom0''s speak the "XEN" arch, so could using cpu specific flags like this cause problems? what are other xen+gentoo users using

I am having a xen 2.0.7 server crash, This is the first message i managed to get out of syslog, I didnt think it was related untill when i did ''xm shutdown mls'' the dom0 also goes down that i thought it might be related, I will see if not running gkrellmd in domU''s will help with this Message from syslogd@michiganlivestock.com at Fri Nov 11 17:25:02 2005 ...

Hi, I have built a Red Hat 7.2 system kernel 2.4.7-10 using pxelinux from syslinux-2.09 I have then installed the latest Broadcom BCM5700 Linux Driver Version 7.1.22 as per the README.TXT:- rpm -ivh bcm5700-7.1.22-1.src.rpm cd /usr/src/redhat rpm -bb SPECS/bcm5700.spec rpm -ivh --force RPMS/i386/bcm5700-7.1.22-1.i386.rpm The system reboots fine and is using the new driver I now want to update my

Hi, i just want to know whether there are any alternative file systems for HP SFS. I heard that there is Cluster Gateway from Polyserve. Can anybody plz help me in finding more abt this Cluster Gateway. Thanks and Regards, Ashok Bharat -----Original Message----- From: lustre-discuss-bounces at lists.lustre.org on behalf of lustre-discuss-request at lists.lustre.org Sent: Tue 2/12/2008 3:18 AM

I'm running RHEL/CentOS 3 (2.4.21-47.ELsmp) with Broadcom NIC and BASP bonding (bcm5700 v8.1.55) When network traffic is ~12MB/sec (== ~120K packets/sec) CPU system load is growing up to 50-60%%. Yes, I know, that small packets (~100 bytes/packet) are not optimal for network throughput, but in the short-middle term I cannot re-write our application. Are there any network driver / kernel

Hi... I have a strange problem with my Private NIC Channel, here is my environment details : 1. RHAS 2.1 with kernel 2.4.9-e.27 Enterprise 2. OCFS version : 2.4.9-e-enterprise-1.0.9-6 3. Oracle RDBMS : 9.2.0.4 RAC with 5 Nodes 4. Private NIC Channel: Broadcom BCM5700 My Private NIC channel is down intermittently, we just tracing the root of the problems by identify each product installed the

Hi, I have a problem with networking under XEN, I hope someone can help me out. The problem is that after some time (a day or so) the network suddenly stops working, and it takes some waiting and/or a reboot to fix this. I''m running XEN 3.1.0-rc7 on an Intel Core2 (x86_64). It''s a Fedora 8 system, with the following packages: o Xen version 3.1.0-rc7-2950.fc8

Is there instructions on how to use the patchless client on RHEL4 ? For version 1.6.2 We would prefer a rpm, but we are not scared of doing a build if needed. Brock Palen Center for Advanced Computing brockp at umich.edu (734)936-1985

I would like to add a driver (bcm5700.ko) in modules.cgz. but there are several files modules.cgz . In which files I do have to add the drivers: - initrd.img ? - stage2.img ? - netstg2.img ? - hdstg2.img ? Denis -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.centos.org/pipermail/centos/attachments/20070118/6af6402d/attachment.html>

Should I be seeing messages like: Dec 14 12:06:59 nyx170 kernel: Lustre: MGS: haven''t heard from client dadccfac-8610-06e7-9c02-90e552694947 (at 141.212.30.185 at tcp) in 234 seconds. I think it''s dead, and I am evicting it. when the client was shut down cleanly? and the lustre file system is mounted via /etc/fstab ? The file system (i would hope) would be unmounted

$ uname -a Linux xxx 2.6.9-22.0.1.ELsmp #1 SMP Thu Oct 27 13:14:25 CDT 2005 i686 i686 i386 GNU/Linux $ lsmod | grep "[^0-9]0" | grepbyhand ip_gre 16865 0 3c59x 41213 0 tg3 85061 0 It is very likely an error that these have usage counts of 0. (the interfaces are very much up, and a GRE tunnel is configured) Seeing as the same thing

Thanks again to Jeff for his time answering our questions about Gluster. I hope people who are new to Gluster find this useful: http://www.rce-cast.com/Podcast/rce-79-gluster-fs.html If you have questions about the podcast please ping me off list. Thanks! Brock Palen www.umich.edu/~brockp CAEN Advanced Computing brockp at umich.edu (734)936-1985

Tim Post schrieb: > On Sat, 2007-01-06 at 10:35 -0500, Brock Palen wrote: >> Isnt forcedeth just a driver for the nvidia nic''s? What would that >> do with broadcom chip? >> > > Doh, I read one post and replied to another. Apologies. too many open > windows. > >> Also, my system is just a pemtium D it does not have the vmx flag >> in

I am one of the hosts of Research Computing and Engineering (www.rce-cast.com). We host shows on HPC related topics and we have had listener requests for a show on Gluster. Would one or two Gluster devs be willing to take an hour for a phone/skype interview to inform users about Gluster and its unique features? If you have any questions please let me know. Brock Palen www.umich.edu/~brockp

We are the hosts of the High Performance Computing Podcast: www.rce-cast.com We were wondering if LLVM / Clang would want to be a feature on the show? It takes about an hour over Skype/Phone and is focussed on being educational. Feel free to contact me (off list) if you are interested in being on the show. Thanks! Brock Palen www.umich.edu/~brockp CAEN Advanced Computing XSEDE Campus

I host and HPC podcast with Jeff Squyres of Open-MPI fame at http://www.rce-cast.com We have had a request to have Puppet featured on the show soon and we would like this also. Feel free to have a developer or two contact me off list and we can setup a mutual time, it takes about 1 hour and is done over the phone or skype. I hope to hear from you soon! Brock Palen

on our cluster that has been running lustre for about 1 month. I have 1 MDT/MGS and 1 OSS with 2 OST''s. Our cluster uses all Gige and has about 608 nodes 1854 cores. We have allot of jobs that die, and/or go into high IO wait, strace shows processes stuck in fstat(). The big problem is (i think) I would like some feedback on it that of these 608 nodes 209 of them have in dmesg

Hi all, I've just upgraded my home network to a Linksys SRW2024 (with lovely vlan trunking support!) and I'm trying to enable jumbo frames. When I try to set the MTU of an rtl8169 based network card (it's a Netgear RA311), it comes up with an error stating: # ifconfig eth1 mtu 9000 SIOCSIFMTU: Invalid argument The only options it will take are 1500 or below. Does anyone know

Hi all Is it possible to configure jumbo frames (mtu=9000) on a kvm guest using virtio drivers? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com