From christian.masopust at siemens.com Tue Feb 1 11:25:07 2011 From: christian.masopust at siemens.com (Masopust, Christian) Date: Tue, 1 Feb 2011 12:25:07 +0100 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <20110122040654.GA5036@hiwaay.net> References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: any news on this topic? Here I have some RHEL6 systems that now and then crash (appr. all 10 days) and currently I've absolutely no idea why (other RHEL6 run fine on same HW) How exactly can/should I monitor the memory usage? Thanks, christian -----Urspr?ngliche Nachricht----- Von: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list-bounces at redhat.com] Im Auftrag von Chris Adams Gesendet: Samstag, 22. J?nner 2011 05:07 An: Stephen John Smoogen Cc: Red Hat Enterprise Linux 6 (Santiago) discussion mailing-list Betreff: Re: [rhelv6-list] Kernel memory leak? Once upon a time, Stephen John Smoogen said: > > I looked at "slabtop", and the dentry cache is the culprit: > > > > ?OBJS ACTIVE ?USE OBJ SIZE ?SLABS OBJ/SLAB CACHE SIZE NAME > > 3078900 3078900 100% ? ?0.19K 153945 ? ? ? 20 ? ?615780K dentry > > > > Anybody else seeing this? > > What is the box doing? How is it set up (ext3, ext4, ?). I haven't > really looked but would want to check with a system that is set up > similarly. It is running ext4 on LVM on md-raid1. It is running Nagios, Apache, Quagga, Network UPS Tools (monitoring a couple of UPSes), and smstools. Quagga is running OSPF (to learn some routes to some of the Nagios-monitored devices) and BGP (to advertise some routes via BGP from a home-written "bad IP" monitor). The bad-IP monitor uses several perl scripts I wrote, one of which uses the Linux::Inotify2 module to watch a directory that gets log files added and removed for each bad IP. The last few days have been rather busy for my bad IP detector; there are 1292 files in that directory right now for the last 48 hours. I wondered if the single inotify could be a trigger (as that's the only thing really unusual), but stopping that daemon doesn't free the RAM from the dentry cache. This same set of software was running on the old server (running Fedora 7 i386 - yes, I was that far behind). It was different hardware, but the same setup except for ext3 instead of ext4 (still LVM on md-raid1, Nagios, Apache, etc.). The old server's RAM usage had been level for years at 256M RAM; the new server started at about 512M (not unexpected since I switched to x86_64) and has increased in an almost perfectly straight line to just under 1G in 8 days. The only other difference from the old server is that it had SELinux disabled and the new one has SELinux running in permissive mode (still trying to work on a useful policy to allow Nagios to do all the things I need). -- Chris Adams Systems and Network Administrator - HiWAAY Internet Services I don't speak for anybody but myself - that's enough trouble. _______________________________________________ rhelv6-list mailing list rhelv6-list at redhat.com https://www.redhat.com/mailman/listinfo/rhelv6-list From mstevens at imt-systems.com Tue Feb 1 15:45:34 2011 From: mstevens at imt-systems.com (Morten P.D. Stevens) Date: Tue, 1 Feb 2011 16:45:34 +0100 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: <7222A6B8ACA37D4AA1AC37810C43E8A8586690677F@mail.corp.imt-systems.com> Hi, We have exactly the same problem in our test environment. (identical IBM xSeries hardware) Some machines crashing after a few days. Note: We are testing with Scientific Linux 6 kernel 2.6.32-71.14.1.el6.x86_64. (But should make no difference to RHEL6) Particularly striking: the large size of dentry in slabtop. Any ideas? Best regards, Morten > -----Original Message----- > From: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list- > bounces at redhat.com] On Behalf Of Masopust, Christian > Sent: Tuesday, February 01, 2011 12:25 PM > To: 'Red Hat Enterprise Linux 6 (Santiago) discussion mailing-list' > Subject: Re: [rhelv6-list] Kernel memory leak? > > > any news on this topic? > > Here I have some RHEL6 systems that now and then crash (appr. all 10 > days) > and currently I've absolutely no idea why (other RHEL6 run fine on same > HW) > > How exactly can/should I monitor the memory usage? > > Thanks, > christian > > > -----Urspr?ngliche Nachricht----- > Von: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list- > bounces at redhat.com] Im Auftrag von Chris Adams > Gesendet: Samstag, 22. J?nner 2011 05:07 > An: Stephen John Smoogen > Cc: Red Hat Enterprise Linux 6 (Santiago) discussion mailing-list > Betreff: Re: [rhelv6-list] Kernel memory leak? > > Once upon a time, Stephen John Smoogen said: > > > I looked at "slabtop", and the dentry cache is the culprit: > > > > > > ?OBJS ACTIVE ?USE OBJ SIZE ?SLABS OBJ/SLAB CACHE SIZE NAME > > > 3078900 3078900 100% ? ?0.19K 153945 ? ? ? 20 ? ?615780K dentry > > > > > > Anybody else seeing this? > > > > What is the box doing? How is it set up (ext3, ext4, ?). I haven't > > really looked but would want to check with a system that is set up > > similarly. > > It is running ext4 on LVM on md-raid1. It is running Nagios, Apache, > Quagga, Network UPS Tools (monitoring a couple of UPSes), and smstools. > Quagga is running OSPF (to learn some routes to some of the > Nagios-monitored devices) and BGP (to advertise some routes via BGP > from > a home-written "bad IP" monitor). > > The bad-IP monitor uses several perl scripts I wrote, one of which uses > the Linux::Inotify2 module to watch a directory that gets log files > added and removed for each bad IP. The last few days have been rather > busy for my bad IP detector; there are 1292 files in that directory > right now for the last 48 hours. > > I wondered if the single inotify could be a trigger (as that's the only > thing really unusual), but stopping that daemon doesn't free the RAM > from the dentry cache. > > This same set of software was running on the old server (running Fedora > 7 i386 - yes, I was that far behind). It was different hardware, but > the same setup except for ext3 instead of ext4 (still LVM on md-raid1, > Nagios, Apache, etc.). The old server's RAM usage had been level for > years at 256M RAM; the new server started at about 512M (not unexpected > since I switched to x86_64) and has increased in an almost perfectly > straight line to just under 1G in 8 days. > > The only other difference from the old server is that it had SELinux > disabled and the new one has SELinux running in permissive mode (still > trying to work on a useful policy to allow Nagios to do all the things > I > need). > > -- > Chris Adams > Systems and Network Administrator - HiWAAY Internet Services > I don't speak for anybody but myself - that's enough trouble. > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list From cmadams at hiwaay.net Tue Feb 1 16:50:33 2011 From: cmadams at hiwaay.net (Chris Adams) Date: Tue, 1 Feb 2011 10:50:33 -0600 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: <20110201165033.GD4060@hiwaay.net> Once upon a time, Masopust, Christian said: > any news on this topic? > > Here I have some RHEL6 systems that now and then crash (appr. all 10 days) > and currently I've absolutely no idea why (other RHEL6 run fine on same HW) > > How exactly can/should I monitor the memory usage? I run the SNMP daemon and graph memory usage with Cricket (Cacti is probably the more popular way these days). My RHEL 6 server doesn't crash from this. I run a network backup system that does a full backup Friday nights and incrementals Monday-Thursday nights. Not much changes on this server, so the incremental takes only takes about 2 minutes (and doesn't do much other than read the directories). The full backup only takes about 7 minutes, but does read all files on the disk (a little over 2G). This is enough that the dentry cache frees a lot of its memory and the system keeps going. I'm not sure that's going to be sufficient in the long term though; the saw-tooth pattern on my graph is still creeping upwards (the drop in RAM usage each week doesn't go as low as the week before). -- Chris Adams Systems and Network Administrator - HiWAAY Internet Services I don't speak for anybody but myself - that's enough trouble. From kmaxnn at gmail.com Tue Feb 1 08:04:46 2011 From: kmaxnn at gmail.com (kelvin maxnn) Date: Tue, 1 Feb 2011 13:34:46 +0530 Subject: [rhelv6-list] Getting kernel panic using custom initrd image Message-ID: Hi all, sorry for my bad English, in advance. Since long time, I have been making custom initrd files for rhel versions like RHEL4.X , RHEL5.X and now trying the same for RHEL6. Using the original vmlinuz (that comes with RHEL iso) and my custom initrd, I used to do pxe booting. My custom initrd I used to create something like this, pseudo steps:- 1. Create all the necessary / directory hierarchy like /etc, /usr, etc.etc. and copy necessary files in that. 2. /bin/dd if=/dev/zero of=/tmp/initrd.img bs=1k count=165888 (the value of count is based on some custom calculations) 3. sbin/mke2fs -b 1024 -q -m0 -F -O none /tmp/initrd.img 4. /bin/mount -o loop /tmp/initrd.img /mycustominitrd 5. Using /bin/cp --recursive --preserve , I used to copy the custom directory hierarchy to /mycustominitrd. 6. unmount /mycustominitrd. 7. Using "gzip --force" , I used to create initrd image (without using mkinitrd or dracut). This customized initrd image along with original vmlinuz , I use to so pxe booting my machines. Up to RHEL5.X, it worked perfectly. But now, it seems that vmlinuz of RHEL6 expects something different than RHEL5 vmlinuz. When I do pxe booting the machine using RHEL6's vmlinuz and my initrd, it takes vmlinuz and initrd through tftp successfully, but after extracting initrd, it gives me error as follows:- RAMDISK: gzip image found at block 0 List of all partitions: No filesystem could mount root, tried: iso9660 Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(9,1) Pid: 1, comm: swapper Not tainted 2.6.32-71.el6.x86_64 #1 Call Trace: ............ (please see the screenshot for full details....) More surprisingly, if I try to use the same initrd with RHEL5 or RHEL4 vmlinuz it works perfectly. This is the some of the contents of pxelinux.cfg file: KERNEL feb1/vmlinuz APPEND ramdisk_blocksize=1024 ramdisk_size=165888 initrd=feb1/initrd.img IPAPPEND 3 Also, if I replace my initrd with initramfs of RHEL6 , i get following error:- dracut: FATAL: No or empty root= argument dracut: Refusing to continue dracut: FATAL: No or empty root= argument dracut: Refusing to continue Signal caught! Boot has failed, sleeping forever. Am I missing something in pxe.cfg ? Does vmlinuz of RHEL6 expects something more than RHEL5 vmlinuz? Can anybody please guide me on this issue? -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: rhel6_kernel_panic.PNG Type: image/png Size: 435503 bytes Desc: not available URL: From solarflow99 at gmail.com Tue Feb 1 19:24:21 2011 From: solarflow99 at gmail.com (solarflow99) Date: Tue, 1 Feb 2011 14:24:21 -0500 Subject: [rhelv6-list] Getting kernel panic using custom initrd image In-Reply-To: References: Message-ID: doesn't rhel6 use dracut for the initrd now? On Tue, Feb 1, 2011 at 3:04 AM, kelvin maxnn wrote: > Hi all, > sorry for my bad?English, in advance. > > Since long time, ?I have been making custom initrd files for rhel versions > like RHEL4.X , RHEL5.X and now trying the same for RHEL6. > Using the original vmlinuz (that comes with RHEL iso) and my custom initrd, > I used to do pxe booting. > My custom initrd I used to create something like this, pseudo steps:- > 1. Create all the necessary / directory?hierarchy?like /etc, /usr, etc.etc. > and copy necessary files in that. > 2.?/bin/dd if=/dev/zero of=/tmp/initrd.img bs=1k count=165888 (the value of > count is based on some custom calculations) > 3.?sbin/mke2fs -b 1024 -q -m0 -F -O none /tmp/initrd.img > 4.?/bin/mount -o loop /tmp/initrd.img /mycustominitrd > 5. Using /bin/cp --recursive --preserve , I used to copy the custom > directory?hierarchy?to?/mycustominitrd. > 6. unmount?/mycustominitrd. > 7.?Using "gzip --force" ,?I used to create initrd image (without using > mkinitrd or dracut). > This?customized?initrd image along with original vmlinuz , I use to so pxe > booting my machines. > Up to RHEL5.X, it worked perfectly. But now, it seems that vmlinuz of RHEL6 > expects something ?different than RHEL5 vmlinuz. > When I do pxe booting the machine using RHEL6's vmlinuz and my initrd, it > takes vmlinuz and initrd through tftp?successfully, but after extracting > initrd, it gives me error as follows:- > RAMDISK: gzip image found at block 0 > List of all partitions: > No filesystem could mount root, tried: ?iso9660 > Kernel panic - not syncing: VFS: Unable to mount root fs on > unknown-block(9,1) > ?Pid: 1, comm: swapper Not tainted 2.6.32-71.el6.x86_64 #1 > ?Call Trace: > ............ (please see the screenshot for full details....) > > More surprisingly, if I try to use the same initrd with RHEL5 or RHEL4 > vmlinuz it works perfectly. > > This is the some of the contents of pxelinux.cfg file: > KERNEL feb1/vmlinuz > APPEND ramdisk_blocksize=1024 ramdisk_size=165888 initrd=feb1/initrd.img > IPAPPEND 3 > > Also, if I replace my initrd with initramfs of RHEL6 , i get following > error:- > dracut: FATAL: No or empty root= argument > dracut: Refusing to continue > dracut: FATAL: No or empty root= argument > dracut: Refusing to continue > > > Signal caught! > > Boot has failed, sleeping forever. > > Am I missing something in pxe.cfg ? > Does vmlinuz of RHEL6 expects something more than RHEL5 vmlinuz? > Can anybody please guide me on this issue? > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > From pmeyer at themeyerfarm.com Tue Feb 1 20:35:43 2011 From: pmeyer at themeyerfarm.com (Phil Meyer) Date: Tue, 01 Feb 2011 13:35:43 -0700 Subject: [rhelv6-list] Getting kernel panic using custom initrd image In-Reply-To: References: Message-ID: <4D486E9F.1000109@themeyerfarm.com> On 02/01/2011 12:24 PM, solarflow99 wrote: > doesn't rhel6 use dracut for the initrd now? > Yes. You will need to install dracut-network before making the initrd image. Otherwise you won't have the network boot capability in there. After installing dracut-network, all future use of dracut for building the initrd will include the network drivers. That is why its not installed by default. example: # yum install dracut-network ... # cd /boot # dracut newinitrd.img `uname -r` That will give you a usable image for network booting. Just copy the current vmlinuz and the newinitrd.img to your tftpserver. Good Luck! From delhage at gmail.com Tue Feb 1 21:17:17 2011 From: delhage at gmail.com (delhage at gmail.com) Date: Tue, 1 Feb 2011 23:17:17 +0200 Subject: [rhelv6-list] EL6 and T410(nvidia) In-Reply-To: <4D0F1247.6010606@gmail.com> References: <4D0F1247.6010606@gmail.com> Message-ID: On Mon, Dec 20, 2010 at 10:22, Grant Williamson wrote: > All, > ? ?anyone on this list using EL6 on a Lenovo T410, which has an nvidia > graphics card. > I am hearing reports from colleagues that the default nouveau driver just > locks up at GDM. > > Anyone else seeing this issue, or have a bugzilla already open? > I've been seeing this exact issue this week on Lenovo T510 with nVidia. lspci reports "nVidia Corporation GT218 [NVS 3100M] (rev a2)". Installation goes fine, but then it locks up when GDM appears or a few seconds after. I haven't found any bugzilla that looks like this problem. My "solution" so far is to blacklist the nouveau driver using "rdblacklist=nouveau" as a bootparameter and adding "blacklist nouveau" to /etc/modprobe.d/blacklist-nouveau.conf. This seems to fix the problem even though Xorg is eventually loading the nouveau driver anyway, I have no clue why it works. I should mention that I've had one problem with this after several hours of work and that is that the screen freezes, although the machine still works (can login remotely). Adding an /etc/X11/xorg.conf specifying the vesa driver instead does work and prevents the nouveau driver from being loaded at all. Cheers, Lars From smooge at gmail.com Wed Feb 2 04:56:44 2011 From: smooge at gmail.com (Stephen John Smoogen) Date: Tue, 1 Feb 2011 21:56:44 -0700 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <20110122040654.GA5036@hiwaay.net> References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: On Fri, Jan 21, 2011 at 21:06, Chris Adams wrote: > Once upon a time, Stephen John Smoogen said: >> > I looked at "slabtop", and the dentry cache is the culprit: >> > >> > ?OBJS ACTIVE ?USE OBJ SIZE ?SLABS OBJ/SLAB CACHE SIZE NAME >> > 3078900 3078900 100% ? ?0.19K 153945 ? ? ? 20 ? ?615780K dentry >> > >> > Anybody else seeing this? >> >> What is the box doing? How is it set up (ext3, ext4, ?). I haven't >> really looked but would want to check with a system that is set up >> similarly. > > It is running ext4 on LVM on md-raid1. ?It is running Nagios, Apache, > Quagga, Network UPS Tools (monitoring a couple of UPSes), and smstools. > Quagga is running OSPF (to learn some routes to some of the > Nagios-monitored devices) and BGP (to advertise some routes via BGP from > a home-written "bad IP" monitor). > > The bad-IP monitor uses several perl scripts I wrote, one of which uses > the Linux::Inotify2 module to watch a directory that gets log files > added and removed for each bad IP. ?The last few days have been rather > busy for my bad IP detector; there are 1292 files in that directory > right now for the last 48 hours. > > I wondered if the single inotify could be a trigger (as that's the only > thing really unusual), but stopping that daemon doesn't free the RAM > from the dentry cache. My guess is that since the dentry cache is kernel space stopping user space items would not have any effect on it. Actually reading some older slides, dentries are cached heavily to make "user experience" faster and are not freed until memory pressure from the kernel. Also looking at various forums, there seem to be many questions about this from around 2.6.27 so I would guess it is something between 2.6.18+patches and 2.6.32+patches which is changing how things work. -- Stephen J Smoogen. "The core skill of innovators is error recovery, not failure avoidance." Randy Nelson, President of Pixar University. "Let us be kind, one to another, for most of us are fighting a hard battle." -- Ian MacLaren From christian.masopust at siemens.com Wed Feb 2 14:06:00 2011 From: christian.masopust at siemens.com (Masopust, Christian) Date: Wed, 2 Feb 2011 15:06:00 +0100 Subject: [rhelv6-list] switch/"rebrand" RHEL to CentOS (or SF-Linux) or vice-versa? Message-ID: is it possible? which packages to replace? thanks a lot, christian -------------- next part -------------- An HTML attachment was scrubbed... URL: From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 2 14:49:42 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 2 Feb 2011 15:49:42 +0100 Subject: [rhelv6-list] switch/"rebrand" RHEL to CentOS (or SF-Linux) or vice-versa? In-Reply-To: References: Message-ID: <20110202154942.6c578546@python3.es.aed.lan> Masopust, Christian wrote : > is it possible? > > which packages to replace? I regularly modify pre-installed CentOS instances to become RHEL, since many cheap-ish hosting companies offer CentOS pre-installed but don't have RHEL available (or charge separately for it and don't let you use an existing entitlement you have). I haven't done it with RHEL6 yet. Partly because the available CentOS are still 5.x and partly because all my quick tests to go from CentOS 5 to RHEL6 have resulted in unbootable remote systems which I can't debug (hint: it should be possible, the most tricky part is that you need to upgrade to an intermediate rpm build which understands the newer xz payloads). Here are some quick steps for a CentOS5 -> RHEL5 change : yum clean all rpm -e --nodeps centos-release centos-release-notes rm -f /etc/yum.repos.d/CentOS-* # Remove some packages to get closer to a minimal system, YMMV yum remove bind bind-chroot caching-nameserver 'man-pages-??' gamin setools glibc-devel specspo # Remove 32bit compat stuff for x86_64 systems yum remove glibc.i686 libgcc.i386 libaio.i386 # Download and install redhat-release* (adapt path, obviously) cd /tmp wget redhat-release-5Server-5.6.0.3.x86_64.rpm wget redhat-release-notes-5Server-36.x86_64.rpm rpm -Uvh redhat-release-* rm -rf /var/cache/yum/* # Once here you need to point the system to your own RHEL yum repo # or install rhn packages Of course, you end up with still a bunch of CentOS packages, all those which had the same version-release as the current RHEL ones. Also, you get a system for which you should definitely not expect to get any official Red Hat support!! You do however get a system which can update from RHN, possibly more consistent with an existing infrastructure. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.14 0.07 0.09 From Iain.Morrison at mrc-epid.cam.ac.uk Wed Feb 2 19:45:18 2011 From: Iain.Morrison at mrc-epid.cam.ac.uk (Iain Morrison) Date: Wed, 2 Feb 2011 19:45:18 -0000 Subject: [rhelv6-list] Kickstart via NFS In-Reply-To: <20110202154942.6c578546@python3.es.aed.lan> References: <20110202154942.6c578546@python3.es.aed.lan> Message-ID: <7F2C60CEE8B98346BD9FBAE010150522323672@silicon.mrc-epid.cam.ac.uk> Dear list, we have a basic kickstart system that works with RHEL 4 and RHEL 5, but is failing with RHEL 6. We boot the machines from CD/DVD and then type linux ks ksdevice=ethX [where X is the relevant number] The IP is set by DHCP and using filename "/export/n_os/ks/"; next-server 11.22.33.44; where 11.22.33.44 is an NFS server and /export/n_os/ks/ the directory containing kickstart files of the form 22.33.44.55-kickstart When we do this for RHEL 6 we get an error downloading kickstart file message. If we enter the location e.g. nfs:11.22.33.44:/export/n_os/ks/22.33.44.55-kickstart the ks file is read. [Putting the ks file location explicitly on the command line also works] Am I missing something? thanks iain -- Iain Morrison IT Manager MRC Epidemiology Unit Institute of Metabolic Science Box 285, Addenbrooke's Hospital Hills Road Cambridge CB2 0QQ Tel 01223 769200 From carlopmart at gmail.com Thu Feb 3 16:00:33 2011 From: carlopmart at gmail.com (carlopmart) Date: Thu, 03 Feb 2011 17:00:33 +0100 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? Message-ID: <4D4AD121.1020300@gmail.com> Hi all, Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I read RedHat's docs about this, and it seems 2GB ... Really?? Is it possible to limit RAM used by the KVM host as a Xen does?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From pmeyer at themeyerfarm.com Thu Feb 3 17:08:51 2011 From: pmeyer at themeyerfarm.com (Phil Meyer) Date: Thu, 03 Feb 2011 10:08:51 -0700 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: <4D4AD121.1020300@gmail.com> References: <4D4AD121.1020300@gmail.com> Message-ID: <4D4AE123.3060502@themeyerfarm.com> On 02/03/2011 09:00 AM, carlopmart wrote: > Hi all, > > Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I > read RedHat's docs about this, and it seems 2GB ... Really?? > > Is it possible to limit RAM used by the KVM host as a Xen does?? > > Thanks. Xen dom0 is a tiny, raw hypervisor using bits of the Linux kernel to operate hardware. KVM is a kernel module based hypervisor. Thus it uses the full kernel and OS with all that implies. VMs utilizing the KVM hypervisor run as applications in the host application space. This is totally different than Xen. It has great advantages, and a few disadvantages. Using KVM, you have to think of the VMs as fixed sized applications. Want to run 2 4GB KVM VMs? The host needs to have more than 8GB available to it in order to dedicate RAM to the VMs as well as manage its own resources. I am bending the specifics slightly to make the point that KVM based VMs run on top of a full kernel/OS, and that KVM VMs are treated as applications within that space. Hope this helps. Good Luck! From danishka at gmail.com Thu Feb 3 17:22:53 2011 From: danishka at gmail.com (Danishka Navin) Date: Thu, 3 Feb 2011 22:52:53 +0530 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: <4D4AE123.3060502@themeyerfarm.com> References: <4D4AD121.1020300@gmail.com> <4D4AE123.3060502@themeyerfarm.com> Message-ID: On Thu, Feb 3, 2011 at 10:38 PM, Phil Meyer wrote: > On 02/03/2011 09:00 AM, carlopmart wrote: > >> Hi all, >> >> Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I read >> RedHat's docs about this, and it seems 2GB ... Really?? >> >> Is it possible to limit RAM used by the KVM host as a Xen does?? >> >> Thanks. >> > > Xen dom0 is a tiny, raw hypervisor using bits of the Linux kernel to > operate hardware. > > KVM is a kernel module based hypervisor. Thus it uses the full kernel and > OS with all that implies. > > VMs utilizing the KVM hypervisor run as applications in the host > application space. This is totally different than Xen. > > It has great advantages, and a few disadvantages. > > Using KVM, you have to think of the VMs as fixed sized applications. > > Want to run 2 4GB KVM VMs? The host needs to have more than 8GB available > to it in order to dedicate RAM to the VMs as well as manage its own > resources. > > but i have given 3GB for my VM where the base system has 2GB of RAM. i was just testing not for production. Base OS is Fedora 14 and Guest is RHEL6 > I am bending the specifics slightly to make the point that KVM based VMs > run on top of a full kernel/OS, and that KVM VMs are treated as applications > within that space. > > Hope this helps. > > Good Luck! > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > -- Danishka Navin http://danishkanavin.blogspot.com http://twitter.com/danishkanavin http://identi.ca/danishka -------------- next part -------------- An HTML attachment was scrubbed... URL: From carlopmart at gmail.com Thu Feb 3 17:26:51 2011 From: carlopmart at gmail.com (carlopmart) Date: Thu, 03 Feb 2011 18:26:51 +0100 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: <4D4AE123.3060502@themeyerfarm.com> References: <4D4AD121.1020300@gmail.com> <4D4AE123.3060502@themeyerfarm.com> Message-ID: <4D4AE55B.3010405@gmail.com> On 02/03/2011 06:08 PM, Phil Meyer wrote: > On 02/03/2011 09:00 AM, carlopmart wrote: >> Hi all, >> >> Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I read RedHat's >> docs about this, and it seems 2GB ... Really?? >> >> Is it possible to limit RAM used by the KVM host as a Xen does?? >> >> Thanks. > > Xen dom0 is a tiny, raw hypervisor using bits of the Linux kernel to operate hardware. > > KVM is a kernel module based hypervisor. Thus it uses the full kernel and OS with > all that implies. > > VMs utilizing the KVM hypervisor run as applications in the host application space. > This is totally different than Xen. > > It has great advantages, and a few disadvantages. > > Using KVM, you have to think of the VMs as fixed sized applications. > > Want to run 2 4GB KVM VMs? The host needs to have more than 8GB available to it in > order to dedicate RAM to the VMs as well as manage its own resources. > > I am bending the specifics slightly to make the point that KVM based VMs run on top > of a full kernel/OS, and that KVM VMs are treated as applications within that space. > > Hope this helps. > > Good Luck! > Thanks Phil. I will run minimal services on host side (ntp, postfix and kvm). In the example that you explain, does enough with the host had, for example, 8.5 GB of RAM? I don't need HA or Clustering services or to connect to external storage (I will use local disks). Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From pmeyer at themeyerfarm.com Thu Feb 3 18:09:46 2011 From: pmeyer at themeyerfarm.com (Phil Meyer) Date: Thu, 03 Feb 2011 11:09:46 -0700 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: <4D4AE55B.3010405@gmail.com> References: <4D4AD121.1020300@gmail.com> <4D4AE123.3060502@themeyerfarm.com> <4D4AE55B.3010405@gmail.com> Message-ID: <4D4AEF6A.4070306@themeyerfarm.com> On 02/03/2011 10:26 AM, carlopmart wrote: > On 02/03/2011 06:08 PM, Phil Meyer wrote: >> On 02/03/2011 09:00 AM, carlopmart wrote: >>> Hi all, >>> >>> Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I >>> read RedHat's >>> docs about this, and it seems 2GB ... Really?? >>> >>> Is it possible to limit RAM used by the KVM host as a Xen does?? >>> >>> Thanks. >> >> Xen dom0 is a tiny, raw hypervisor using bits of the Linux kernel to >> operate hardware. >> >> KVM is a kernel module based hypervisor. Thus it uses the full kernel >> and OS with >> all that implies. >> >> VMs utilizing the KVM hypervisor run as applications in the host >> application space. >> This is totally different than Xen. >> >> It has great advantages, and a few disadvantages. >> >> Using KVM, you have to think of the VMs as fixed sized applications. >> >> Want to run 2 4GB KVM VMs? The host needs to have more than 8GB >> available to it in >> order to dedicate RAM to the VMs as well as manage its own resources. >> >> I am bending the specifics slightly to make the point that KVM based >> VMs run on top >> of a full kernel/OS, and that KVM VMs are treated as applications >> within that space. >> >> Hope this helps. >> >> Good Luck! >> > > Thanks Phil. I will run minimal services on host side (ntp, postfix > and kvm). In the example that you explain, does enough with the host > had, for example, 8.5 GB of RAM? > > I don't need HA or Clustering services or to connect to external > storage (I will use local disks). > > Thanks. > > Here is a real scenario: I have a well equipped server with 32GB RAM installed. I installed RHEL6 with kvm, libvirt, virt-manager, ksm, etc. I have personally created 60 KVM based VMs running Fedora 14 set to 512MB RAM each on this system. They all functioned within specifications for the hardware they were running on. The host computer showed almost exactly 30GB used by the guests under load, and as low as 17GB when all 60 were initially booted and idle. Therefore, I can conclude that a 32GB system can manage up to 60 512MB KVM based VMs. I/O for those guests is another matter entirely! :) You will drive the host crazy unless you turn of cache for the VMs, and use some kind of fast storage array for all those clients. But that is another plus of using a real host based hypervisor. In this same scenario I could run 4 2GB KVM based VMs, for example, and leave disk cacheing on in the config. This way the host will use its large memory reserve of 24GB as mostly disk cache for my VMs. This can produce results of over 500MB/sec reads in bonnie++ inside the VMs! Dangerous, but fun! Using the host as a disk cache for the VMs can cause a lot of issues on a crash, and prevents live migration of the VM, but man, the VMs disk i/o screams! Good Luck! From carlopmart at gmail.com Thu Feb 3 18:47:58 2011 From: carlopmart at gmail.com (carlopmart) Date: Thu, 03 Feb 2011 19:47:58 +0100 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: <4D4AEF6A.4070306@themeyerfarm.com> References: <4D4AD121.1020300@gmail.com> <4D4AE123.3060502@themeyerfarm.com> <4D4AE55B.3010405@gmail.com> <4D4AEF6A.4070306@themeyerfarm.com> Message-ID: <4D4AF85E.6050500@gmail.com> On 02/03/2011 07:09 PM, Phil Meyer wrote: > On 02/03/2011 10:26 AM, carlopmart wrote: >> On 02/03/2011 06:08 PM, Phil Meyer wrote: >>> On 02/03/2011 09:00 AM, carlopmart wrote: >>>> Hi all, >>>> >>>> Which is the minmal amount of RAM needed for a RHEL6 KVM host?? I read RedHat's >>>> docs about this, and it seems 2GB ... Really?? >>>> >>>> Is it possible to limit RAM used by the KVM host as a Xen does?? >>>> >>>> Thanks. >>> >>> Xen dom0 is a tiny, raw hypervisor using bits of the Linux kernel to operate >>> hardware. >>> >>> KVM is a kernel module based hypervisor. Thus it uses the full kernel and OS with >>> all that implies. >>> >>> VMs utilizing the KVM hypervisor run as applications in the host application space. >>> This is totally different than Xen. >>> >>> It has great advantages, and a few disadvantages. >>> >>> Using KVM, you have to think of the VMs as fixed sized applications. >>> >>> Want to run 2 4GB KVM VMs? The host needs to have more than 8GB available to it in >>> order to dedicate RAM to the VMs as well as manage its own resources. >>> >>> I am bending the specifics slightly to make the point that KVM based VMs run on top >>> of a full kernel/OS, and that KVM VMs are treated as applications within that space. >>> >>> Hope this helps. >>> >>> Good Luck! >>> >> >> Thanks Phil. I will run minimal services on host side (ntp, postfix and kvm). In >> the example that you explain, does enough with the host had, for example, 8.5 GB >> of RAM? >> >> I don't need HA or Clustering services or to connect to external storage (I will >> use local disks). >> >> Thanks. >> >> > > Here is a real scenario: > > I have a well equipped server with 32GB RAM installed. > > I installed RHEL6 with kvm, libvirt, virt-manager, ksm, etc. > > I have personally created 60 KVM based VMs running Fedora 14 set to 512MB RAM each > on this system. > > They all functioned within specifications for the hardware they were running on. > > The host computer showed almost exactly 30GB used by the guests under load, and as > low as 17GB when all 60 were initially booted and idle. > > Therefore, I can conclude that a 32GB system can manage up to 60 512MB KVM based VMs. > > I/O for those guests is another matter entirely! :) > > You will drive the host crazy unless you turn of cache for the VMs, and use some > kind of fast storage array for all those clients. > > But that is another plus of using a real host based hypervisor. > > In this same scenario I could run 4 2GB KVM based VMs, for example, and leave disk > cacheing on in the config. This way the host will use its large memory reserve of > 24GB as mostly disk cache for my VMs. This can produce results of over 500MB/sec > reads in bonnie++ inside the VMs! > > Dangerous, but fun! > > Using the host as a disk cache for the VMs can cause a lot of issues on a crash, and > prevents live migration of the VM, but man, the VMs disk i/o screams! > > Good Luck! > Many thanks Phil. -- CL Martinez carlopmart {at} gmail {d0t} com From paolo.campegiani at gmail.com Fri Feb 4 08:50:25 2011 From: paolo.campegiani at gmail.com (Paolo Campegiani) Date: Fri, 4 Feb 2011 09:50:25 +0100 Subject: [rhelv6-list] Minimal RAM for a RHEL6 KVM host? In-Reply-To: References: <4D4AD121.1020300@gmail.com> <4D4AE123.3060502@themeyerfarm.com> Message-ID: >> Want to run 2 4GB KVM VMs? ?The host needs to have more than 8GB available >> to it in order to dedicate RAM to the VMs as well as manage its own >> resources. Not necessarily, and usually it's not so. It depends a lot on the workload on the guest systems and on the ability of the virtualization layer to handle a memory pressure scenario. > but i have given 3GB for my VM where the base system has 2GB of RAM. > i was just testing not for production. > > Base OS is Fedora 14 and Guest is RHEL6 This guest memory overcommitting is possible by: - using swap memory of the host system; - using memory overcommit on the host system; - using ksm on the host system. (possibly others, this is not intended to be an exhaustive list). You can check that there isn't any magic involved in this by running a small program on the guest that allocates big chunks of memory and fill them with data from /dev/urandom (to defeat the ksm merging mechanism) (of course without releasing the memory area after having filled it). From carlopmart at gmail.com Fri Feb 4 08:50:37 2011 From: carlopmart at gmail.com (carlopmart) Date: Fri, 04 Feb 2011 09:50:37 +0100 Subject: [rhelv6-list] How can I create sparse block devices as a logical volumes with lvm Message-ID: <4D4BBDDD.5070707@gmail.com> Hi all, Is it possible to create sparse block devices with LVM (logical volumes not volume groups) on RHEL6 like ZFS does?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From john.haxby at gmail.com Fri Feb 4 11:37:01 2011 From: john.haxby at gmail.com (John Haxby) Date: Fri, 4 Feb 2011 11:37:01 +0000 Subject: [rhelv6-list] How can I create sparse block devices as a logical volumes with lvm In-Reply-To: <4D4BBDDD.5070707@gmail.com> References: <4D4BBDDD.5070707@gmail.com> Message-ID: On 4 February 2011 08:50, carlopmart wrote: > Hi all, > > Is it possible to create sparse block devices with LVM (logical volumes > not volume groups) on RHEL6 like ZFS does?? > > Do you mean something like this (from the lvcreate man page): "lvcreate --virtualsize 1T --size 100M --snapshot --name sparse vg1" creates a sparse device named /dev/vg1/sparse of size 1TB with space for just under 100MB of actual data on it. jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From carlopmart at gmail.com Fri Feb 4 11:45:29 2011 From: carlopmart at gmail.com (carlopmart) Date: Fri, 04 Feb 2011 12:45:29 +0100 Subject: [rhelv6-list] How can I create sparse block devices as a logical volumes with lvm In-Reply-To: References: <4D4BBDDD.5070707@gmail.com> Message-ID: <4D4BE6D9.7030308@gmail.com> On 02/04/2011 12:37 PM, John Haxby wrote: > > > On 4 February 2011 08:50, carlopmart > wrote: > > Hi all, > > Is it possible to create sparse block devices with LVM (logical volumes not > volume groups) on RHEL6 like ZFS does?? > > > Do you mean something like this (from the lvcreate man page): > > "lvcreate --virtualsize 1T --size 100M --snapshot --name sparse vg1" > creates a sparse device named /dev/vg1/sparse of size 1TB with space > for just under 100MB of actual data on it. > > > jch > > Yes, but If I understand good, these options are only valid for snapshots ... I am trying to do a sparse logical volume without using snapshots ... -- CL Martinez carlopmart {at} gmail {d0t} com From john.haxby at gmail.com Fri Feb 4 12:05:45 2011 From: john.haxby at gmail.com (John Haxby) Date: Fri, 4 Feb 2011 12:05:45 +0000 Subject: [rhelv6-list] How can I create sparse block devices as a logical volumes with lvm In-Reply-To: <4D4BE6D9.7030308@gmail.com> References: <4D4BBDDD.5070707@gmail.com> <4D4BE6D9.7030308@gmail.com> Message-ID: On 4 February 2011 11:45, carlopmart wrote: > On 02/04/2011 12:37 PM, John Haxby wrote: > >> >> Do you mean something like this (from the lvcreate man page): >> >> "lvcreate --virtualsize 1T --size 100M --snapshot --name sparse vg1" >> creates a sparse device named /dev/vg1/sparse of size 1TB with >> space >> for just under 100MB of actual data on it. >> >> >> jch >> >> >> > Yes, but If I understand good, these options are only valid for snapshots > ... I am trying to do a sparse logical volume without using snapshots ... > > Hmmm. So you want to create a sparse logical volume using something other than the mechanism for creating sparse logical volumes? Quoting the man page again: --virtualsize VirtualSize Create a sparse device of the given size (in MB by default) using a snapshot. Anything written to the device will be returned when reading from it. Reading from other areas of the device will return blocks of zeros. It is implemented by creat? ing a hidden virtual device of the requested size using the zero target. A suffix of _vorigin is used for this device. Actually, I think the "--snapshot" on the man page example I previously quoted is either redundant or wrong, this command works just fine: lvcreate --virtualsize 1T --size 100M --name sparse vg1 Now it doesn't mention the s-word are you happy? jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From jfranz at freerun.com Fri Feb 4 12:47:26 2011 From: jfranz at freerun.com (Benjamin Franz) Date: Fri, 04 Feb 2011 04:47:26 -0800 Subject: [rhelv6-list] How can I create sparse block devices as a logical volumes with lvm In-Reply-To: References: <4D4BBDDD.5070707@gmail.com> Message-ID: <4D4BF55E.1090605@freerun.com> On 02/04/2011 03:37 AM, John Haxby wrote: > > > On 4 February 2011 08:50, carlopmart > wrote: > > Hi all, > > Is it possible to create sparse block devices with LVM (logical > volumes not volume groups) on RHEL6 like ZFS does?? > > > Do you mean something like this (from the lvcreate man page): > > "lvcreate --virtualsize 1T --size 100M --snapshot --name > sparse vg1" > creates a sparse device named /dev/vg1/sparse of size 1TB > with space > for just under 100MB of actual data on it. > Ok. Cool feature from a technical perspective....But...What is the use case? I'm trying to wrap my head over why I would want to overcommit a logical volume's usable storage. I'm assuming there is some use case where it makes sense, but I'm not seeing it. The closest I can come is that you might manually resize the lv's 'real' size if you ever got close to filling it up. Enlighten me? -- Benjamin Franz -------------- next part -------------- An HTML attachment was scrubbed... URL: From digital at pghfilmmakers.org Fri Feb 4 16:07:11 2011 From: digital at pghfilmmakers.org (Pittsburgh Filmmakers Institute) Date: Fri, 4 Feb 2011 11:07:11 -0500 Subject: [rhelv6-list] openSwan, IPSEC, and RHEL6 Message-ID: <201102041107.11637.digital@pghfilmmakers.org> Is anyone aware of some really good, proper openSwan configuration documentation or tutorials? it seems RHEL6's only officially supported option for VPN server IS openSwan, but so far there doesn't seem to be any RHEL documentation about it aside from a list of significant IPSEC files. Things I've already tried: openswan wiki, rhel6 security guide, red hat official support (it seems a lot of the reps aren't terribly familiar with openSwan yet either). The must-have's: network-to-network connectivity between two campus buildings both running RHEL The nice-to-have's: road warrior connectivity for the random worker out in the field or working from home Other alternatives I could consider: installing a (non-supported) rpm of openVPN and going that route, regressing to RHEL5 and using a supported rpm of openVPN. --seth From carlopmart at gmail.com Sat Feb 5 16:44:02 2011 From: carlopmart at gmail.com (carlopmart) Date: Sat, 05 Feb 2011 17:44:02 +0100 Subject: [rhelv6-list] Whereis gphoto2 package for RHEL6? Message-ID: <4D4D7E52.30908@gmail.com> Hi all, I am trying to extract some photos from my Canon camera and I can't. With RHEL5.x I had used gphoto2 package, but I can't see it on RHEL6 list packages ... Where is?? Or how can I extract photos from a camera on RHEL6?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From amyagi at gmail.com Sat Feb 5 17:03:12 2011 From: amyagi at gmail.com (Akemi Yagi) Date: Sat, 5 Feb 2011 09:03:12 -0800 Subject: [rhelv6-list] Whereis gphoto2 package for RHEL6? In-Reply-To: <4D4D7E52.30908@gmail.com> References: <4D4D7E52.30908@gmail.com> Message-ID: On Sat, Feb 5, 2011 at 8:44 AM, carlopmart wrote: > Hi all, > > ?I am trying to extract some photos from my Canon camera and I can't. With > RHEL5.x I had used gphoto2 package, but I can't see it on RHEL6 list > packages ... Where is?? Or how can I extract photos from a camera on RHEL6?? I have not tried this myself but I believe libgphoto2 is the package you need in RHEL6. Akemi From carlopmart at gmail.com Sat Feb 5 17:09:54 2011 From: carlopmart at gmail.com (carlopmart) Date: Sat, 05 Feb 2011 18:09:54 +0100 Subject: [rhelv6-list] Whereis gphoto2 package for RHEL6? In-Reply-To: <4D4D7E52.30908@gmail.com> References: <4D4D7E52.30908@gmail.com> Message-ID: <4D4D8462.8030105@gmail.com> On 02/05/2011 05:44 PM, carlopmart wrote: > Hi all, > > I am trying to extract some photos from my Canon camera and I can't. With RHEL5.x I > had used gphoto2 package, but I can't see it on RHEL6 list packages ... Where is?? > Or how can I extract photos from a camera on RHEL6?? > > Thanks. > OOpps .. Sorry. Package that I need is gthumb ... and I have found it. -- CL Martinez carlopmart {at} gmail {d0t} com From tinnisg at gmail.com Mon Feb 7 00:41:45 2011 From: tinnisg at gmail.com (Tinnis G) Date: Sun, 6 Feb 2011 19:41:45 -0500 Subject: [rhelv6-list] Facing issue to register the server Message-ID: Hi, I have a test server where rhel5.2 86_64 bit OS was installed. Recently I have upgraded to rhel6. The previous OS was registered with redhat. That is removed . Now when I am trying to execute rhn_register , the following errors are coming We can't contact the Red Hat Network Server, double check the location provided - is https://xmlrpc.rhn.redhat.com/XCMLRPC' correct ? Make sure that the network connection on this system is operational. Apparently everything looks ok , i mean all the necessary files are there... The iptables has been disabled. Where I am missing. If anybody has any idea please share.. Thanks in advance -------------- next part -------------- An HTML attachment was scrubbed... URL: From marco.shaw at gmail.com Mon Feb 7 00:52:34 2011 From: marco.shaw at gmail.com (Marco Shaw) Date: Sun, 6 Feb 2011 20:52:34 -0400 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: References: Message-ID: Are you sure about this part "XCMLRPC"? It should be "XMLRPC". Marco On Sun, Feb 6, 2011 at 8:41 PM, Tinnis G wrote: > Hi, > > I have a test server where rhel5.2 86_64 bit OS was installed.? Recently I > have upgraded to rhel6. > The previous OS was registered with redhat.? That is removed . > > Now when I am trying to execute rhn_register , the following errors? are > coming > > We can't contact the Red Hat Network Server, double check the location > provided - is > https://xmlrpc.rhn.redhat.com/XCMLRPC' correct ? > Make sure that the network connection on this system is operational. > > Apparently everything looks ok , i mean all the necessary files are there... > The iptables has been disabled. > > Where I am missing. If anybody has any idea please share.. > > Thanks in advance > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > -- *Microsoft MVP - Windows PowerShell https://mvp.support.microsoft.com/profile/Marco.Shaw *Co-Author - Sams Windows PowerShell Unleashed 2nd Edition *Blog - http://marcoshaw.blogspot.com From tinnisg at gmail.com Mon Feb 7 03:22:34 2011 From: tinnisg at gmail.com (Tinnis G) Date: Sun, 6 Feb 2011 22:22:34 -0500 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: References: Message-ID: Sorry, it is my mistake , it is "XMLRPC". Thanks again. On Sun, Feb 6, 2011 at 7:52 PM, Marco Shaw wrote: > Are you sure about this part "XCMLRPC"? It should be "XMLRPC". > > Marco > > On Sun, Feb 6, 2011 at 8:41 PM, Tinnis G wrote: > > Hi, > > > > I have a test server where rhel5.2 86_64 bit OS was installed. Recently > I > > have upgraded to rhel6. > > The previous OS was registered with redhat. That is removed . > > > > Now when I am trying to execute rhn_register , the following errors are > > coming > > > > We can't contact the Red Hat Network Server, double check the location > > provided - is > > https://xmlrpc.rhn.redhat.com/XCMLRPC' correct ? > > Make sure that the network connection on this system is operational. > > > > Apparently everything looks ok , i mean all the necessary files are > there... > > The iptables has been disabled. > > > > Where I am missing. If anybody has any idea please share.. > > > > Thanks in advance > > > > _______________________________________________ > > rhelv6-list mailing list > > rhelv6-list at redhat.com > > https://www.redhat.com/mailman/listinfo/rhelv6-list > > > > > > > > -- > *Microsoft MVP - Windows PowerShell > https://mvp.support.microsoft.com/profile/Marco.Shaw > *Co-Author - Sams Windows PowerShell Unleashed 2nd Edition > *Blog - http://marcoshaw.blogspot.com > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > -------------- next part -------------- An HTML attachment was scrubbed... URL: From robert at b2b4linux.nl Mon Feb 7 06:51:31 2011 From: robert at b2b4linux.nl (Robert Maessen) Date: Mon, 07 Feb 2011 07:51:31 +0100 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: References: Message-ID: <1297061491.2638.1.camel@vision.b2b4linux.nl> Maybe an obvious question, but is your DNS working? Robert On Sun, 2011-02-06 at 22:22 -0500, Tinnis G wrote: > Sorry, it is my mistake , it is "XMLRPC". > > Thanks again. > > > On Sun, Feb 6, 2011 at 7:52 PM, Marco Shaw > wrote: > > Are you sure about this part "XCMLRPC"? It should be > "XMLRPC". > > Marco > > > > On Sun, Feb 6, 2011 at 8:41 PM, Tinnis G > wrote: > > Hi, > > > > I have a test server where rhel5.2 86_64 bit OS was > installed. Recently I > > have upgraded to rhel6. > > The previous OS was registered with redhat. That is > removed . > > > > Now when I am trying to execute rhn_register , the following > errors are > > coming > > > > We can't contact the Red Hat Network Server, double check > the location > > provided - is > > https://xmlrpc.rhn.redhat.com/XCMLRPC' correct ? > > Make sure that the network connection on this system is > operational. > > > > Apparently everything looks ok , i mean all the necessary > files are there... > > The iptables has been disabled. > > > > Where I am missing. If anybody has any idea please share.. > > > > Thanks in advance > > > > > > _______________________________________________ > > rhelv6-list mailing list > > rhelv6-list at redhat.com > > https://www.redhat.com/mailman/listinfo/rhelv6-list > > > > > > > > -- > *Microsoft MVP - Windows PowerShell > https://mvp.support.microsoft.com/profile/Marco.Shaw > *Co-Author - Sams Windows PowerShell Unleashed 2nd Edition > *Blog - http://marcoshaw.blogspot.com > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list -------------- next part -------------- An HTML attachment was scrubbed... URL: From jean-yves at lenhof.eu.org Mon Feb 7 07:28:11 2011 From: jean-yves at lenhof.eu.org (Jean-Yves LENHOF) Date: Mon, 07 Feb 2011 08:28:11 +0100 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: <1297061491.2638.1.camel@vision.b2b4linux.nl> References: <1297061491.2638.1.camel@vision.b2b4linux.nl> Message-ID: <4D4F9F0B.7090701@lenhof.eu.org> Le 07/02/2011 07:51, Robert Maessen a ?crit : > Maybe an obvious question, but is your DNS working? > > Robert verify if you have not a clock screw, this can make a problem... regards, From carlopmart at gmail.com Mon Feb 7 09:42:57 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 07 Feb 2011 10:42:57 +0100 Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 Message-ID: <4D4FBEA1.9040302@gmail.com> Hi all, I have a very strange problem. I have installed a virtual RHEL6 machine under an ESXi host using paravirtual disk devices. I need to assign static names to these disk but I can't. I have read the following guide form RedHat's knowledge: https://access.redhat.com/kb/docs/DOC-45628, but it doesn't works for vmware paravirtual devices. For example: [root at hobbiton by-path]# ls -al total 0 drwxr-xr-x 2 root root 200 Feb 7 10:29 . drwxr-xr-x 4 root root 80 Feb 7 10:08 .. lrwxrwxrwx 1 root root 9 Feb 7 10:08 pci-0000:03:00.0-scsi-0:0:0:0 -> ../../sda lrwxrwxrwx 1 root root 10 Feb 7 10:08 pci-0000:03:00.0-scsi-0:0:0:0-part1 -> ../../sda1 lrwxrwxrwx 1 root root 10 Feb 7 10:08 pci-0000:03:00.0-scsi-0:0:0:0-part2 -> ../../sda2 lrwxrwxrwx 1 root root 9 Feb 7 10:29 pci-0000:04:00.0-scsi-0:0:0:0 -> ../../sdb lrwxrwxrwx 1 root root 10 Feb 7 10:29 pci-0000:04:00.0-scsi-0:0:0:0-part1 -> ../../sdb1 lrwxrwxrwx 1 root root 9 Feb 7 10:08 pci-0000:04:00.0-scsi-0:0:1:0 -> ../../sdc lrwxrwxrwx 1 root root 9 Feb 7 10:08 pci-0000:0c:00.0-scsi-0:0:0:0 -> ../../sdd lrwxrwxrwx 1 root root 9 Feb 7 10:08 pci-0000:14:00.0-scsi-0:0:0:0 -> ../../sde I need to rename sdb, sdc, sdd and sde disks. But when I execute scsi_id command, it doesn't returns any result: [root at hobbiton by-path]# scsi_id --whitelisted --replace-whitespace --device=/dev/sdb [root at hobbiton by-path]# If I use lsilogic driver on these disks instead pvscsi drivers, all works as a expected but I loose performance. My question is: how can I assign static device names with udev rules using by-path or ID_PATH strings?? I need to create lvm volumes with these disks. Many thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From jean-yves at lenhof.eu.org Mon Feb 7 12:02:54 2011 From: jean-yves at lenhof.eu.org (Jean-Yves LENHOF) Date: Mon, 7 Feb 2011 13:02:54 +0100 (CET) Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 In-Reply-To: <4D4FBEA1.9040302@gmail.com> References: <4D4FBEA1.9040302@gmail.com> Message-ID: <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> Le Lun 7 f?vrier 2011 10:42, carlopmart a ?crit : > Hi all, > My question is: how can I assign static device names with udev rules > using by-path > or ID_PATH strings?? > > I need to create lvm volumes with these disks. If you'll be using using LVM, why do you want a specific name for your disk ? What is the point ? regards, From carlopmart at gmail.com Mon Feb 7 12:26:20 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 07 Feb 2011 13:26:20 +0100 Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 In-Reply-To: <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> References: <4D4FBEA1.9040302@gmail.com> <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> Message-ID: <4D4FE4EC.3040806@gmail.com> On 02/07/2011 01:02 PM, Jean-Yves LENHOF wrote: > Le Lun 7 f?vrier 2011 10:42, carlopmart a ?crit : >> Hi all, > > > >> My question is: how can I assign static device names with udev rules >> using by-path >> or ID_PATH strings?? >> >> I need to create lvm volumes with these disks. > > If you'll be using using LVM, why do you want a specific name for your > disk ? What is the point ? > > regards, > I have configured 4 disks: - sdb: adapter 1, lun 0 - sdc: adapter 1, lun 1 - sdd: adapter 2, lun 0 - sde: adapter 3, lun 0 If I need to add another virtual harddisk on adapter 1 with lun 2, all disks names after sdc will change. I need to avoid that. -- CL Martinez carlopmart {at} gmail {d0t} com From jean-yves at lenhof.eu.org Mon Feb 7 12:44:26 2011 From: jean-yves at lenhof.eu.org (Jean-Yves LENHOF) Date: Mon, 7 Feb 2011 13:44:26 +0100 (CET) Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 In-Reply-To: <4D4FE4EC.3040806@gmail.com> References: <4D4FBEA1.9040302@gmail.com> <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> <4D4FE4EC.3040806@gmail.com> Message-ID: <31da0ccb02bc0d156ebe8f26659824f1.squirrel@wwws.aidi.eu.org> Le Lun 7 f?vrier 2011 13:26, carlopmart a ?crit : > On 02/07/2011 01:02 PM, Jean-Yves LENHOF wrote: >> Le Lun 7 f?vrier 2011 10:42, carlopmart a ?crit : >>> Hi all, >> >> >> >>> My question is: how can I assign static device names with udev rules >>> using by-path >>> or ID_PATH strings?? >>> >>> I need to create lvm volumes with these disks. >> >> If you'll be using using LVM, why do you want a specific name for your >> disk ? What is the point ? >> >> regards, >> > > I have configured 4 disks: > > - sdb: adapter 1, lun 0 > - sdc: adapter 1, lun 1 > - sdd: adapter 2, lun 0 > - sde: adapter 3, lun 0 > > If I need to add another virtual harddisk on adapter 1 with lun 2, all > disks names > after sdc will change. I need to avoid that. And ? There must be a specific command to see the mapping between the LUN and discs if you want this information at a specific moment. But for me, but disk name can be changed and lvm will find all your logical volume without a hitch during the vgchange command at boot time... All the needed information are stored in the private part of the lvm partition of the disk. Regards, From john.haxby at gmail.com Mon Feb 7 12:47:38 2011 From: john.haxby at gmail.com (John Haxby) Date: Mon, 7 Feb 2011 12:47:38 +0000 Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 In-Reply-To: <4D4FE4EC.3040806@gmail.com> References: <4D4FBEA1.9040302@gmail.com> <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> <4D4FE4EC.3040806@gmail.com> Message-ID: On 7 February 2011 12:26, carlopmart wrote: > > I have configured 4 disks: > > - sdb: adapter 1, lun 0 > - sdc: adapter 1, lun 1 > - sdd: adapter 2, lun 0 > - sde: adapter 3, lun 0 > > If I need to add another virtual harddisk on adapter 1 with lun 2, all > disks names after sdc will change. I need to avoid that. You don't say why though. You can simply side-step the renaming issue by using, for example, /dev/disk/by-* -- pick the one that's most useful to you; I'm guessing you want /dev/disk/by-id. jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From carlopmart at gmail.com Mon Feb 7 13:43:44 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 07 Feb 2011 14:43:44 +0100 Subject: [rhelv6-list] Renaming scsi disks with udev rules on RHEL6 In-Reply-To: <31da0ccb02bc0d156ebe8f26659824f1.squirrel@wwws.aidi.eu.org> References: <4D4FBEA1.9040302@gmail.com> <3b9c4128eb0d31820feaf4d69916f2dd.squirrel@wwws.aidi.eu.org> <4D4FE4EC.3040806@gmail.com> <31da0ccb02bc0d156ebe8f26659824f1.squirrel@wwws.aidi.eu.org> Message-ID: <4D4FF710.5000902@gmail.com> On 02/07/2011 01:44 PM, Jean-Yves LENHOF wrote: > Le Lun 7 f?vrier 2011 13:26, carlopmart a ?crit : >> On 02/07/2011 01:02 PM, Jean-Yves LENHOF wrote: >>> Le Lun 7 f?vrier 2011 10:42, carlopmart a ?crit : >>>> Hi all, >>> >>> >>> >>>> My question is: how can I assign static device names with udev rules >>>> using by-path >>>> or ID_PATH strings?? >>>> >>>> I need to create lvm volumes with these disks. >>> >>> If you'll be using using LVM, why do you want a specific name for your >>> disk ? What is the point ? >>> >>> regards, >>> >> >> I have configured 4 disks: >> >> - sdb: adapter 1, lun 0 >> - sdc: adapter 1, lun 1 >> - sdd: adapter 2, lun 0 >> - sde: adapter 3, lun 0 >> >> If I need to add another virtual harddisk on adapter 1 with lun 2, all >> disks names >> after sdc will change. I need to avoid that. > > And ? > > There must be a specific command to see the mapping between the LUN and > discs if you want this information at a specific moment. > > But for me, but disk name can be changed and lvm will find all your > logical volume without a hitch during the vgchange command at boot time... > All the needed information are stored in the private part of the lvm > partition of the disk. > > Regards, > > Ouch .. You are right Jean-Yves. Many thanks fr your help. -- CL Martinez carlopmart {at} gmail {d0t} com From tinnisg at gmail.com Mon Feb 7 14:48:43 2011 From: tinnisg at gmail.com (Tinnis G) Date: Mon, 7 Feb 2011 09:48:43 -0500 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: <1297061491.2638.1.camel@vision.b2b4linux.nl> References: <1297061491.2638.1.camel@vision.b2b4linux.nl> Message-ID: Yes, everything is fine. The /etc/hosts, /etc/resolv.conf, are being configured , the way I generally configure on the other servers. No problem with name resolution, ssh, ftp snmp ... thanks again On Mon, Feb 7, 2011 at 1:51 AM, Robert Maessen wrote: > Maybe an obvious question, but is your DNS working? > > Robert > > > On Sun, 2011-02-06 at 22:22 -0500, Tinnis G wrote: > > Sorry, it is my mistake , it is "XMLRPC". > > Thanks again. > > On Sun, Feb 6, 2011 at 7:52 PM, Marco Shaw wrote: > > Are you sure about this part "XCMLRPC"? It should be "XMLRPC". > > Marco > > > > On Sun, Feb 6, 2011 at 8:41 PM, Tinnis G wrote: > > Hi, > > > > I have a test server where rhel5.2 86_64 bit OS was installed. Recently > I > > have upgraded to rhel6. > > The previous OS was registered with redhat. That is removed . > > > > Now when I am trying to execute rhn_register , the following errors are > > coming > > > > We can't contact the Red Hat Network Server, double check the location > > provided - is > > https://xmlrpc.rhn.redhat.com/XCMLRPC' correct ? > > Make sure that the network connection on this system is operational. > > > > Apparently everything looks ok , i mean all the necessary files are > there... > > The iptables has been disabled. > > > > Where I am missing. If anybody has any idea please share.. > > > > Thanks in advance > > > > > _______________________________________________ > > rhelv6-list mailing list > > rhelv6-list at redhat.com > > https://www.redhat.com/mailman/listinfo/rhelv6-list > > > > > > > > -- > *Microsoft MVP - Windows PowerShell > https://mvp.support.microsoft.com/profile/Marco.Shaw > *Co-Author - Sams Windows PowerShell Unleashed 2nd Edition > *Blog - http://marcoshaw.blogspot.com > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > > _______________________________________________ > rhelv6-list mailing listrhelv6-list at redhat.comhttps://www.redhat.com/mailman/listinfo/rhelv6-list > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From franchu.garcia at gmail.com Mon Feb 7 20:43:05 2011 From: franchu.garcia at gmail.com (Fran Garcia) Date: Mon, 7 Feb 2011 21:43:05 +0100 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: References: <1297061491.2638.1.camel@vision.b2b4linux.nl> Message-ID: On Mon, Feb 7, 2011 at 15:48, Tinnis G wrote: > Yes, everything is fine. The /etc/hosts, /etc/resolv.conf, are being > configured , the way I generally configure on the > other servers. No problem with name resolution, ssh, ftp snmp ... > > I might be stating the obvious, but if you're using RH Satellite, you need version 5.4 minimum. RHEL6 won't work on anything previous that release. If you're using hosred RHN it will be a network/connectivity issue :-) Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From KCollins at chevron.com Mon Feb 7 23:42:40 2011 From: KCollins at chevron.com (Collins, Kevin [BEELINE]) Date: Mon, 7 Feb 2011 23:42:40 +0000 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: <4D4F9F0B.7090701@lenhof.eu.org> References: <1297061491.2638.1.camel@vision.b2b4linux.nl> <4D4F9F0B.7090701@lenhof.eu.org> Message-ID: <6F56410FBED1FC41BCA804E16F594B0B019DDD@chvpkw8xmbx01.chvpk.chevrontexaco.net> I assume you mean 'clock skew'? -----Original Message----- From: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list-bounces at redhat.com] On Behalf Of Jean-Yves LENHOF Sent: Sunday, February 06, 2011 11:28 PM To: rhelv6-list at redhat.com Subject: Re: [rhelv6-list] Facing issue to register the server Le 07/02/2011 07:51, Robert Maessen a ?crit : > Maybe an obvious question, but is your DNS working? > > Robert verify if you have not a clock screw, this can make a problem... regards, _______________________________________________ rhelv6-list mailing list rhelv6-list at redhat.com https://www.redhat.com/mailman/listinfo/rhelv6-list From jean-yves at lenhof.eu.org Tue Feb 8 00:19:30 2011 From: jean-yves at lenhof.eu.org (Jean-Yves LENHOF) Date: Tue, 08 Feb 2011 01:19:30 +0100 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: <6F56410FBED1FC41BCA804E16F594B0B019DDD@chvpkw8xmbx01.chvpk.chevrontexaco.net> References: <1297061491.2638.1.camel@vision.b2b4linux.nl> <4D4F9F0B.7090701@lenhof.eu.org> <6F56410FBED1FC41BCA804E16F594B0B019DDD@chvpkw8xmbx01.chvpk.chevrontexaco.net> Message-ID: <4D508C12.4010606@lenhof.eu.org> Le 08/02/2011 00:42, Collins, Kevin [BEELINE] a ?crit : > I assume you mean 'clock skew'? yep my bad From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 9 09:47:21 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 9 Feb 2011 10:47:21 +0100 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) Message-ID: <20110209104721.0a971fee@python3.es.aed.lan> Hi, I just got this isolated RHEL6 server, with hardware I've never used before. It's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap SATA disks (WD10EADS-00P8B0) using software RAID-1. The system is unusable : A quick strace of various processes shows very long waits during many different I/O related calls such as open, read, pread, stat, fdatasync... I've tried failing every other partition in order to be sure to use only one disk for a while, with the same results. Running RAID rebuilds sometimes peaks at 60-100MB/s but very rarely. Right now for instance, the main 900GB RAID-1 is rebuilding from sda2 to sdb2 and it's reporting 400kB/s with system load of over 2 (the server is idle apart from this rebuild). While this is going on, even smartctl on sdb takes time to respond. When no rebuild is in progress, the problems are the same. I'm using the latest released 2.6.32-71.14.1.el6.x86_64 kernel, and here is the detail of the SATA controller(s) : 00:1f.2 IDE interface: Intel Corporation 82801JI (ICH10 Family) 4 port SATA IDE Controller #1 Subsystem: Super Micro Computer Inc Device 0009 Kernel driver in use: ata_piix Kernel modules: ata_generic, pata_acpi, ata_piix 00:1f.5 IDE interface: Intel Corporation 82801JI (ICH10 Family) 2 port SATA IDE Controller #2 Subsystem: Super Micro Computer Inc Device 0009 Kernel driver in use: ata_piix Kernel modules: ata_generic, pata_acpi, ata_piix No errors reported by the kernel apart from some processes stuck for more than 120s before the last reboot. Has anyone seen disk I/O issues like this with RHEL6 and similar hardware? Matthias PS: A possibly important detail is that this is a server provided by ovh.com which has been offering RHEL6 for a while but still has it marked as "BETA" (it's not the RHEL6 beta, it's how they consider their offering of RHEL6), which might indicate known problems between their typical hardware and RHEL6. I didn't find any clues with a quick bugzilla search... -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.00 0.15 0.26 From wolfy at nobugconsulting.ro Wed Feb 9 09:59:31 2011 From: wolfy at nobugconsulting.ro (Manuel Wolfshant) Date: Wed, 09 Feb 2011 11:59:31 +0200 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <20110209104721.0a971fee@python3.es.aed.lan> References: <20110209104721.0a971fee@python3.es.aed.lan> Message-ID: <4D526583.8000503@nobugconsulting.ro> On 02/09/2011 11:47 AM, Matthias Saou wrote: > Hi, > > I just got this isolated RHEL6 server, with hardware I've never used > before. It's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap > SATA disks (WD10EADS-00P8B0) using software RAID-1. > > The system is unusable : A quick strace of various processes shows very > long waits during many different I/O related calls such as open, read, > pread, stat, fdatasync... > > I've tried failing every other partition in order to be sure to use only > one disk for a while, with the same results. Running RAID rebuilds > sometimes peaks at 60-100MB/s but very rarely. Right now for instance, > the main 900GB RAID-1 is rebuilding from sda2 to sdb2 and it's > reporting 400kB/s with system load of over 2 (the server is idle apart > from this rebuild). While this is going on, even smartctl on sdb takes > time to respond. When no rebuild is in progress, the problems are the > same. > > I'm using the latest released 2.6.32-71.14.1.el6.x86_64 kernel, and > here is the detail of the SATA controller(s) : > > 00:1f.2 IDE interface: Intel Corporation 82801JI (ICH10 Family) 4 port > SATA IDE Controller #1 Subsystem: Super Micro Computer Inc Device 0009 > Kernel driver in use: ata_piix > Kernel modules: ata_generic, pata_acpi, ata_piix > 00:1f.5 IDE interface: Intel Corporation 82801JI (ICH10 Family) 2 port > SATA IDE Controller #2 Subsystem: Super Micro Computer Inc Device 0009 > Kernel driver in use: ata_piix > Kernel modules: ata_generic, pata_acpi, ata_piix Can you please check that the drives are seen by the system via the SCSI ( as opposed to older PATA ) protocol ? The BIOS must be configured to avoid legacy/compatible mode by all means ( and use AHCI instead ) From christian.masopust at siemens.com Wed Feb 9 11:00:33 2011 From: christian.masopust at siemens.com (Masopust, Christian) Date: Wed, 9 Feb 2011 12:00:33 +0100 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... Message-ID: Hi all, some of my RHEL6-systems are facing a kernel panic from time to time. 2 of them are huge HP's (DL585 G7) with 48cores and 128GB, one of them is an older Primergy RX300S2 (4 cores, 8GB). Some other systems (also Primergies) run fine all the time... some other facts: - all filesystems ext4 - nfs4 enabled - 3 bonding devices, each having 2 physical devices - 2 of the bonding devices configured for jumbo frames (MTU=9000) Here's the console-log from one of the HP's: ------------[ cut here ]------------ kernel BUG at fs/inode.c:1333! invalid opcode: 0000 [#1] SMP last sysfs file: /sys/devices/system/cpu/cpu47/cache/index2/shared_cpu_map CPU 4 Modules linked in: iptable_filter ip_tables nfs fscache fuse nfsd nfs_acl auth_rpcgss exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler lockd sunrpc bonding ipv6 dm_mirror dm_region_hash dm_log uinput power_meter hwmon bnx2 amd64_edac_mod edac_core edac_mce_amd i2c_piix4 sg h pilo nx_nic(U) ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic pata_atiixp ahci hpsa(U) radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core dm_mod [last unloaded: freq_table] Modules linked in: iptable_filter ip_tables nfs fscache fuse nfsd nfs_acl auth_rpcgss exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler lockd sunrpc bonding ipv6 dm_mirror dm_region_hash dm_log uinput power_meter hwmon bnx2 amd64_edac_mod edac_core edac_mce_amd i2c_piix4 sg h pilo nx_nic(U) ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic pata_atiixp ahci hpsa(U) radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core dm_mod [last unloaded: freq_table] Pid: 3393, comm: lockd Tainted: G W ---------------- 2.6.32-71.14.1.el6.x86_64 #1 ProLiant DL585 G7 RIP: 0010:[] [] iput+0x69/0x70 RSP: 0018:ffff88082b86fce0 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8802fc8616c8 RCX: 000000000000c60e RDX: ffff88202e13a901 RSI: ffffffffa0341de0 RDI: ffff8802fc8616c8 RBP: ffff88082b86fcf0 R08: 000000000002ac45 R09: 0000000000000000 R10: 000000000000000f R11: 0000000000000000 R12: ffff880227b49c00 R13: ffffffffa034e060 R14: ffff88202e13a940 R15: 00000000fffffff5 FS: 00007fac6a0247c0(0000) GS:ffff88002c240000(0000) knlGS:00000000f77916c0 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 00007fac6a048000 CR3: 0000000c2da36000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff4ff0 DR7: 0000000000000400 Process lockd (pid: 3393, threadinfo ffff88082b86e000, task ffff88082d9ab4e0) Stack: ffff88082b86fd40 ffff8802fc861680 ffff88082b86fd10 ffffffff813fdbf1 <0> ffff880227b49c00 ffff880227b49c00 ffff88082b86fd30 ffffffffa03351f8 <0> ffff88082b86fd30 ffff880227b49c10 ffff88082b86fd60 ffffffffa0341e2c Call Trace: [] sock_release+0x71/0x90 [] svc_sock_free+0x48/0x70 [sunrpc] [] svc_xprt_free+0x4c/0x70 [sunrpc] [] ? svc_xprt_free+0x0/0x70 [sunrpc] [] kref_put+0x37/0x70 [] svc_xprt_put+0x19/0x20 [sunrpc] [] svc_xprt_release+0xc1/0xe0 [sunrpc] [] svc_recv+0x2ed/0x830 [sunrpc] [] ? default_wake_function+0x0/0x20 [] lockd+0xc1/0x230 [lockd] [] ? lockd+0x0/0x230 [lockd] [] kthread+0x96/0xa0 [] child_rip+0xa/0x20 [] ? kthread+0x0/0xa0 [] ? child_rip+0x0/0x20 Code: 38 48 c7 c0 f0 7c 18 81 48 85 d2 74 12 48 8b 42 20 48 c7 c2 f0 7c 18 81 48 85 c0 48 0f 44 c2 48 89 df ff d0 48 83 c4 08 5b c9 c3 <0f> 0b eb fe 0f 1f 00 55 48 89 e5 41 55 41 54 53 48 83 ec 08 0f RIP [] iput+0x69/0x70 RSP ?Mounting proc filesystem Mounting sysfs filesystem Creating /dev Creating initial device nodes Free memory/Total memory (free %): 456164 / 495584 ( 92.0457 ) Loading jbd2.ko module Loading mbcache.ko module Loading ext4.ko module Loading crc-t10dif.ko module Loading sd_mod.ko module Loading ata_generic.ko module Loading exportfs.ko module Loading autofs4.ko module Loading ipmi_msghandler.ko module Loading sunrpc.ko module Loading ipv6.ko module Loading uinput.ko module Loading hwmon.ko module Loading bnx2.ko module Loading edac_core.ko module Loading edac_mce_amd.ko module Loading sg.ko module Loading hpilo.ko module Loading nx_nic.ko module Loading cdrom.ko module Loading pata_acpi.ko module Loading pata_atiixp.ko module Loading ahci.ko module Loading hpsa.ko module hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out Loading i2c-core.ko module Loading dm-mod.ko module Loading nfs_acl.ko module Loading auth_rpcgss.ko module Loading ipmi_devintf.ko module Loading ipmi_si.ko module Loading lockd.ko module Loadingpower_meter ACPI000D:00: Ignoring unsafe software power cap! bonding.ko module Loading dm-log.ko module Loading power_meter.ko module Loading amd64_edac_mod.ko module Loading i2c-piix4.ko module Loading sr_mod.ko module Loading drm.ko module Loading i2c-algo-bit.ko module Loading nfsd.ko module Loading dm-region-hash.ko module Loading ttm.ko module Loading drm_kms_helper.ko module Loading dm-mirror.ko module Loading radeon.ko module Waiting for required block device discovery Waiting for 8 sdd-like device(s)...Found Creating Block Devices Creating block device loop0 Creating block device loop1 Creating block device loop2 Creating block device loop3 Creating block device loop4 Creating block device loop5 Creating block device loop6 Creating block device loop7 Creating block device ram0 Creating block device ram1 Creating block device ram10 Creating block device ram11 Creating block device ram12 Creating block device ram13 Creating block device ram14 Creating block device ram15 Creating block device ram2 Creating block device ram3 Creating block device ram4 Creating block device ram5 Creating block device ram6 Creating block device ram7 Creating block device ram8 Creating block device ram9 Creating block device sda Creating block device sdb Creating block device sdc Creating block device sdd Creating block device sr0 mdadm: No arrays found in config file or automatically Free memory/Total memory (free %): 432796 / 495584 ( 87.3305 ) Saving to the local filesystem /dev/sdd1 e2fsck 1.41.12 (17-May-2010) Homes: recovering journal Homes: clean, 9073003/164782080 files, 387571383/659105347 blocks Free memory/Total memory (free %): 427248 / 495584 ( 86.211 ) Copying data : [ 2 %] Copying data : [100 %] Saving core complete Restarting system. Backtrace from crash-dump utility shows: GNU gdb (GDB) 7.0 Copyright (C) 2009 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-unknown-linux-gnu"... KERNEL: /usr/lib/debug/lib/modules/2.6.32-71.14.1.el6.x86_64/vmlinux DUMPFILE: ./vmcore [PARTIAL DUMP] CPUS: 48 DATE: Wed Feb 9 09:30:52 2011 UPTIME: 14 days, 13:57:19 LOAD AVERAGE: 3.65, 3.39, 3.25 TASKS: 1663 NODENAME: hydra.sie.siemens.at RELEASE: 2.6.32-71.14.1.el6.x86_64 VERSION: #1 SMP Wed Jan 5 17:01:01 EST 2011 MACHINE: x86_64 (2095 Mhz) MEMORY: 128 GB PANIC: "kernel BUG at fs/inode.c:1333!" PID: 3393 COMMAND: "lockd" TASK: ffff88082d9ab4e0 [THREAD_INFO: ffff88082b86e000] CPU: 4 STATE: TASK_RUNNING (PANIC) crash> bt PID: 3393 TASK: ffff88082d9ab4e0 CPU: 4 COMMAND: "lockd" #0 [ffff88082b86f9a0] machine_kexec at ffffffff8103695b #1 [ffff88082b86fa00] crash_kexec at ffffffff810b9068 #2 [ffff88082b86fad0] oops_end at ffffffff814cc6e0 #3 [ffff88082b86fb00] die at ffffffff8101733b #4 [ffff88082b86fb30] do_trap at ffffffff814cbfb4 #5 [ffff88082b86fb90] do_invalid_op at ffffffff81014ee5 #6 [ffff88082b86fc30] invalid_op at ffffffff81013f5b [exception RIP: iput+105] RIP: ffffffff81186bf9 RSP: ffff88082b86fce0 RFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8802fc8616c8 RCX: 000000000000c60e RDX: ffff88202e13a901 RSI: ffffffffa0341de0 RDI: ffff8802fc8616c8 RBP: ffff88082b86fcf0 R8: 000000000002ac45 R9: 0000000000000000 R10: 000000000000000f R11: 0000000000000000 R12: ffff880227b49c00 R13: ffffffffa034e060 R14: ffff88202e13a940 R15: 00000000fffffff5 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 #7 [ffff88082b86fcf8] sock_release at ffffffff813fdbf1 #8 [ffff88082b86fd18] svc_sock_free at ffffffffa03351f8 #9 [ffff88082b86fd38] svc_xprt_free at ffffffffa0341e2c #10 [ffff88082b86fd68] kref_put at ffffffff8125cb97 #11 [ffff88082b86fd88] svc_xprt_put at ffffffffa0340f29 #12 [ffff88082b86fd98] svc_xprt_release at ffffffffa0341191 #13 [ffff88082b86fdc8] svc_recv at ffffffffa03415bd #14 [ffff88082b86fe58] lockd at ffffffffa02f6291 #15 [ffff88082b86fee8] kthread at ffffffff81091a76 #16 [ffff88082b86ff48] kernel_thread at ffffffff810141ca crash> any idea? any hint? what else can i do to find the reason for these panics? how to solve it? thanks a lot, christian -------------- next part -------------- An HTML attachment was scrubbed... URL: From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 9 11:05:17 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 9 Feb 2011 12:05:17 +0100 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <4D526583.8000503@nobugconsulting.ro> References: <20110209104721.0a971fee@python3.es.aed.lan> <4D526583.8000503@nobugconsulting.ro> Message-ID: <20110209120517.5bdec31a@python3.es.aed.lan> Manuel Wolfshant wrote : > Can you please check that the drives are seen by the system via the SCSI > ( as opposed to older PATA ) protocol ? The BIOS must be configured to > avoid legacy/compatible mode by all means ( and use AHCI instead ) I'm guessing you're referring to a problem I've indeed seen on some similar hardware which had 82801I (ICH9 Family) and where the default I saw with RHEL5 was indeed to have the SATA drives appear as hda and hdc and resulted in very poor performance. That problem is solved by passing hda=noprobe and hdc=noprobe to the kernel. It's not what I'm seeing now, where both disks are sda and sdb, and the problem isn't that all disk transfers are slow, it's that I/O calls seem to stall very often for multiple seconds. Right now, the RAID-1 rebuild from sda2 to sdb2 is at 40MB/s sustained with all 8 CPUs (Quad HT) at 0.0% wait and a load of 1. This is what I would expect from an idle system. But moments ago, and for hours, it was nearly stalled with always at least 1 CPU at 90%+ wait and all of those I/O pauses for everything. I'm really confused, and still not even sure wether it's a hardware or software problem in the first place. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.06 0.03 0.07 From john.haxby at gmail.com Wed Feb 9 12:58:36 2011 From: john.haxby at gmail.com (John Haxby) Date: Wed, 9 Feb 2011 12:58:36 +0000 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <20110209120517.5bdec31a@python3.es.aed.lan> References: <20110209104721.0a971fee@python3.es.aed.lan> <4D526583.8000503@nobugconsulting.ro> <20110209120517.5bdec31a@python3.es.aed.lan> Message-ID: On 9 February 2011 11:05, Matthias Saou < thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net> wrote: > Manuel Wolfshant wrote : > > > Can you please check that the drives are seen by the system via the SCSI > > ( as opposed to older PATA ) protocol ? The BIOS must be configured to > > avoid legacy/compatible mode by all means ( and use AHCI instead ) > > I'm guessing you're referring to a problem I've indeed seen on some > similar hardware which had 82801I (ICH9 Family) and where the default I > saw with RHEL5 was indeed to have the SATA drives appear as hda and hdc > and resulted in very poor performance. That problem is solved by > passing hda=noprobe and hdc=noprobe to the kernel. It's not what I'm > seeing now, where both disks are sda and sdb, and the problem isn't > that all disk transfers are slow, it's that I/O calls seem to stall > very often for multiple seconds. > > The disks will be sda and sdb regardless of whether they're connected as PATA or SATA. If I was to guess from the information you provided earlier, I'd say they're connected as PATA. This being the case, you need to dive into the BIOS and turn legacy access (or whatever it's called) off and turn AHCI on. You can tell which you're using, though, by looking at the output of dmesg | egrep 'scsi|ata[0-9]': scsi0 : ahci scsi1 : ahci scsi2 : ahci ata1: SATA max UDMA/133 abar m2048 at 0xf6dfb800 port 0xf6dfb900 irq 43 ata2: DUMMY ata3: SATA max UDMA/133 abar m2048 at 0xf6dfb800 port 0xf6dfba00 irq 43 ata_piix 0000:00:1f.1: version 2.13 ata_piix 0000:00:1f.1: PCI INT A -> GSI 16 (level, low) -> IRQ 16 ata_piix 0000:00:1f.1: setting latency timer to 64 scsi3 : ata_piix scsi4 : ata_piix ata4: PATA max UDMA/100 cmd 0x1f0 ctl 0x3f6 bmdma 0x6fa0 irq 14 ata5: PATA max UDMA/100 cmd 0x170 ctl 0x376 bmdma 0x6fa8 irq 15 ata5: port disabled. ignoring. ata4.00: ATA-6: ST980815A, 3.ADE, max UDMA/100 ata4.00: 156301488 sectors, multi 8: LBA48 ata4.00: configured for UDMA/100 ata3: SATA link down (SStatus 0 SControl 300) ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300) ata1.00: ATA-8: FUJITSU MHY2120BH, 0085000B, max UDMA/100 ata1.00: 234441648 sectors, multi 8: LBA48 NCQ (depth 31/32), AA ata1.00: configured for UDMA/100 That's actually my latop which has a SATA disk (ST980815A) and a PATA disk (MHY2120BH). The former is (if you work back through the info) attached to an ahci controller, the latter to an ata_piix controller. I think your controllers were using ata_piix drivers which suggests that the BIOS is being "kind" and presenting you with backwards compatible and slow drivers. I think my BIOS has all kinds of dire warnings about how your system won't boot if your switch to AHCI after you've installed an operating system. By which, of course, it means Windows which can't cope with the change. jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 9 16:35:16 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 9 Feb 2011 17:35:16 +0100 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: References: <20110209104721.0a971fee@python3.es.aed.lan> <4D526583.8000503@nobugconsulting.ro> <20110209120517.5bdec31a@python3.es.aed.lan> Message-ID: <20110209173516.30f55eeb@python3.es.aed.lan> John Haxby wrote : > The disks will be sda and sdb regardless of whether they're connected as > PATA or SATA. If I was to guess from the information you provided earlier, > I'd say they're connected as PATA. This being the case, you need to dive > into the BIOS and turn legacy access (or whatever it's called) off and turn > AHCI on. > > You can tell which you're using, though, by looking at the output of dmesg | > egrep 'scsi|ata[0-9]': > > scsi0 : ahci > scsi1 : ahci > scsi2 : ahci > ata1: SATA max UDMA/133 abar m2048 at 0xf6dfb800 port 0xf6dfb900 irq 43 > ata2: DUMMY > ata3: SATA max UDMA/133 abar m2048 at 0xf6dfb800 port 0xf6dfba00 irq 43 > ata_piix 0000:00:1f.1: version 2.13 > ata_piix 0000:00:1f.1: PCI INT A -> GSI 16 (level, low) -> IRQ 16 > ata_piix 0000:00:1f.1: setting latency timer to 64 > scsi3 : ata_piix > scsi4 : ata_piix > ata4: PATA max UDMA/100 cmd 0x1f0 ctl 0x3f6 bmdma 0x6fa0 irq 14 > ata5: PATA max UDMA/100 cmd 0x170 ctl 0x376 bmdma 0x6fa8 irq 15 > ata5: port disabled. ignoring. > ata4.00: ATA-6: ST980815A, 3.ADE, max UDMA/100 > ata4.00: 156301488 sectors, multi 8: LBA48 > ata4.00: configured for UDMA/100 > ata3: SATA link down (SStatus 0 SControl 300) > ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300) > ata1.00: ATA-8: FUJITSU MHY2120BH, 0085000B, max UDMA/100 > ata1.00: 234441648 sectors, multi 8: LBA48 NCQ (depth 31/32), AA > ata1.00: configured for UDMA/100 > > > That's actually my latop which has a SATA disk (ST980815A) and a PATA disk > (MHY2120BH). The former is (if you work back through the info) attached to > an ahci controller, the latter to an ata_piix controller. I think your > controllers were using ata_piix drivers which suggests that the BIOS is > being "kind" and presenting you with backwards compatible and slow drivers. > > I think my BIOS has all kinds of dire warnings about how your system won't > boot if your switch to AHCI after you've installed an operating system. By > which, of course, it means Windows which can't cope with the change. Thanks for all of those details. I mostly know about all this already, but not only do I not have access to the BIOS (this is one of those servers where hardware is managed by the hosting company on a large scale, any OS is reinstallable remotely at any time, so it makes sense for them to set BIOS defaults which will work with any OS), but the issue I'm seeing is not just I/O being slow. It's I/O being blocked entirely. A simple "ls" can sometimes take tens of seconds when the system is idle. At other times, I do manage to see disks peak at multi-10MB/s read or write speeds. Weird... I was just really hoping someone had already seen this problem with RHEL6, since otherwise it might indicate a hardware issue. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.01 0.02 0.06 From hescominsoon at emmanuelcomputerconsulting.com Thu Feb 10 00:19:35 2011 From: hescominsoon at emmanuelcomputerconsulting.com (William Warren) Date: Wed, 09 Feb 2011 19:19:35 -0500 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <20110209104721.0a971fee@python3.es.aed.lan> References: <20110209104721.0a971fee@python3.es.aed.lan> Message-ID: <4D532F17.8020109@emmanuelcomputerconsulting.com> On 2/9/2011 4:47 AM, Matthias Saou wrote: > Hi, > > I just got this isolated RHEL6 server, with hardware I've never used > before. It's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap > SATA disks (WD10EADS-00P8B0) using software RAID-1. > > The system is unusable : A quick strace of various processes shows very > long waits during many different I/O related calls such as open, read, > pread, stat, fdatasync... > > I've tried failing every other partition in order to be sure to use only > one disk for a while, with the same results. Running RAID rebuilds > sometimes peaks at 60-100MB/s but very rarely. Right now for instance, > the main 900GB RAID-1 is rebuilding from sda2 to sdb2 and it's > reporting 400kB/s with system load of over 2 (the server is idle apart > from this rebuild). While this is going on, even smartctl on sdb takes > time to respond. When no rebuild is in progress, the problems are the > same. > > I'm using the latest released 2.6.32-71.14.1.el6.x86_64 kernel, and > here is the detail of the SATA controller(s) : > > 00:1f.2 IDE interface: Intel Corporation 82801JI (ICH10 Family) 4 port > SATA IDE Controller #1 Subsystem: Super Micro Computer Inc Device 0009 > Kernel driver in use: ata_piix > Kernel modules: ata_generic, pata_acpi, ata_piix > 00:1f.5 IDE interface: Intel Corporation 82801JI (ICH10 Family) 2 port > SATA IDE Controller #2 Subsystem: Super Micro Computer Inc Device 0009 > Kernel driver in use: ata_piix > Kernel modules: ata_generic, pata_acpi, ata_piix > > No errors reported by the kernel apart from some processes stuck for > more than 120s before the last reboot. Has anyone seen disk I/O issues > like this with RHEL6 and similar hardware? > > Matthias > > PS: A possibly important detail is that this is a server provided by > ovh.com which has been offering RHEL6 for a while but still has it > marked as "BETA" (it's not the RHEL6 beta, it's how they consider their > offering of RHEL6), which might indicate known problems between their > typical hardware and RHEL6. I didn't find any clues with a quick > bugzilla search... > are you using the bios raid or MD raid? If you are using biod raid you are running everything effectively pio mode and it's going to suck. Put the bios in ahci(not raid mode) and then use Linux raid from inside the partitioning scheme(which is md raid) i bet performance will be much better. From incoming-redhat at rjl.com Thu Feb 10 05:55:53 2011 From: incoming-redhat at rjl.com (Nataraj) Date: Wed, 09 Feb 2011 21:55:53 -0800 Subject: [rhelv6-list] openSwan, IPSEC, and RHEL6 In-Reply-To: <201102041107.11637.digital@pghfilmmakers.org> References: <201102041107.11637.digital@pghfilmmakers.org> Message-ID: <4D537DE9.6010902@rjl.com> On 02/04/2011 08:07 AM, Pittsburgh Filmmakers Institute wrote: > Is anyone aware of some really good, proper openSwan configuration > documentation or tutorials? it seems RHEL6's only officially supported option > for VPN server IS openSwan, but so far there doesn't seem to be any RHEL > documentation about it aside from a list of significant IPSEC files. > > Things I've already tried: openswan wiki, rhel6 security guide, red hat official > support (it seems a lot of the reps aren't terribly familiar with openSwan yet > either). > > The must-have's: network-to-network connectivity between two campus buildings > both running RHEL > > The nice-to-have's: road warrior connectivity for the random worker out in the > field or working from home > > Other alternatives I could consider: installing a (non-supported) rpm of > openVPN and going that route, regressing to RHEL5 and using a supported rpm of > openVPN. > > --seth > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list I've used openvpn from the epel repo quite sucessfully. Nataraj From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Feb 10 10:39:44 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 10 Feb 2011 11:39:44 +0100 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <4D532F17.8020109@emmanuelcomputerconsulting.com> References: <20110209104721.0a971fee@python3.es.aed.lan> <4D532F17.8020109@emmanuelcomputerconsulting.com> Message-ID: <20110210113944.61a1ccdd@python3.es.aed.lan> William Warren wrote : > On 2/9/2011 4:47 AM, Matthias Saou wrote: > > Hi, > > > > I just got this isolated RHEL6 server, with hardware I've never used > > before. It's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap > > SATA disks (WD10EADS-00P8B0) using software RAID-1. > > > > The system is unusable : A quick strace of various processes shows very > > long waits during many different I/O related calls such as open, read, > > pread, stat, fdatasync... > > > > I've tried failing every other partition in order to be sure to use only > > one disk for a while, with the same results. Running RAID rebuilds > > sometimes peaks at 60-100MB/s but very rarely. Right now for instance, > > the main 900GB RAID-1 is rebuilding from sda2 to sdb2 and it's > > reporting 400kB/s with system load of over 2 (the server is idle apart > > from this rebuild). While this is going on, even smartctl on sdb takes > > time to respond. When no rebuild is in progress, the problems are the > > same. > > > > I'm using the latest released 2.6.32-71.14.1.el6.x86_64 kernel, and > > here is the detail of the SATA controller(s) : > > > > 00:1f.2 IDE interface: Intel Corporation 82801JI (ICH10 Family) 4 port > > SATA IDE Controller #1 Subsystem: Super Micro Computer Inc Device 0009 > > Kernel driver in use: ata_piix > > Kernel modules: ata_generic, pata_acpi, ata_piix > > 00:1f.5 IDE interface: Intel Corporation 82801JI (ICH10 Family) 2 port > > SATA IDE Controller #2 Subsystem: Super Micro Computer Inc Device 0009 > > Kernel driver in use: ata_piix > > Kernel modules: ata_generic, pata_acpi, ata_piix > > > > No errors reported by the kernel apart from some processes stuck for > > more than 120s before the last reboot. Has anyone seen disk I/O issues > > like this with RHEL6 and similar hardware? > > > > Matthias > > > > PS: A possibly important detail is that this is a server provided by > > ovh.com which has been offering RHEL6 for a while but still has it > > marked as "BETA" (it's not the RHEL6 beta, it's how they consider their > > offering of RHEL6), which might indicate known problems between their > > typical hardware and RHEL6. I didn't find any clues with a quick > > bugzilla search... > > > are you using the bios raid or MD raid? If you are using biod raid you > are running everything effectively pio mode and it's going to suck. Put > the bios in ahci(not raid mode) and then use Linux raid from inside the > partitioning scheme(which is md raid) i bet performance will be much better. I probably forgot to detail, but it's indeed MD RAID I'm using. No one in their right mind would use one of those cheap motherboard pseudo RAID, would they? :-) Anyway, I've reinstalled the server with RHEL5 and it's running fine now. Definitely something going on between this particular hardware and RHEL6.0... since I intend to use it only as a KVM hypervisor, and that RHEL6 still doesn't support the new cool stuff that would make a difference (like KSM), I'll stick with RHEL5, even though all of the virtual servers running under it will be RHEL6. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.07 0.07 0.08 From carlopmart at gmail.com Thu Feb 10 10:55:13 2011 From: carlopmart at gmail.com (carlopmart) Date: Thu, 10 Feb 2011 11:55:13 +0100 Subject: [rhelv6-list] nfs process doesn't shutdown clean Message-ID: <4D53C411.5070502@gmail.com> Hi all, I have a very strange problem with two RHEL6 servers that acts as a nfs servers. When i do "shutdown -h now" or "reboot", nfs process never shutdowns cleanly (in fact, "nfs stop" never appears and nfs startup script is configured to do it). Every time, nfsd process locks lvm volumes. Somebody knows if is this a bug?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From solarflow99 at gmail.com Thu Feb 10 14:13:13 2011 From: solarflow99 at gmail.com (solarflow99) Date: Thu, 10 Feb 2011 09:13:13 -0500 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <20110210113944.61a1ccdd@python3.es.aed.lan> References: <20110209104721.0a971fee@python3.es.aed.lan> <4D532F17.8020109@emmanuelcomputerconsulting.com> <20110210113944.61a1ccdd@python3.es.aed.lan> Message-ID: you know, I wonder if there is a chance it could be the CPU bug that was just reported? I had a PC with an i5 that had very poor performance, once reverted to an older release it wasn't as noticeable. On Thu, Feb 10, 2011 at 5:39 AM, Matthias Saou wrote: > William Warren wrote : > >> On 2/9/2011 4:47 AM, Matthias Saou wrote: >> > Hi, >> > >> > I just got this isolated RHEL6 server, with hardware I've never used >> > before. It's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap >> > SATA disks (WD10EADS-00P8B0) using software RAID-1. >> > >> > The system is unusable : A quick strace of various processes shows very >> > long waits during many different I/O related calls such as open, read, >> > pread, stat, fdatasync... >> > >> > I've tried failing every other partition in order to be sure to use only >> > one disk for a while, with the same results. Running RAID rebuilds >> > sometimes peaks at 60-100MB/s but very rarely. Right now for instance, >> > the main 900GB RAID-1 is rebuilding from sda2 to sdb2 and it's >> > reporting 400kB/s with system load of over 2 (the server is idle apart >> > from this rebuild). While this is going on, even smartctl on sdb takes >> > time to respond. When no rebuild is in progress, the problems are the >> > same. >> > >> > I'm using the latest released 2.6.32-71.14.1.el6.x86_64 kernel, and >> > here is the detail of the SATA controller(s) : >> > >> > 00:1f.2 IDE interface: Intel Corporation 82801JI (ICH10 Family) 4 port >> > SATA IDE Controller #1 Subsystem: Super Micro Computer Inc Device 0009 >> > ? ? Kernel driver in use: ata_piix >> > ? ? Kernel modules: ata_generic, pata_acpi, ata_piix >> > 00:1f.5 IDE interface: Intel Corporation 82801JI (ICH10 Family) 2 port >> > SATA IDE Controller #2 Subsystem: Super Micro Computer Inc Device 0009 >> > ? ? Kernel driver in use: ata_piix >> > ? ? Kernel modules: ata_generic, pata_acpi, ata_piix >> > >> > No errors reported by the kernel apart from some processes stuck for >> > more than 120s before the last reboot. Has anyone seen disk I/O issues >> > like this with RHEL6 and similar hardware? >> > >> > Matthias >> > >> > PS: A possibly important detail is that this is a server provided by >> > ovh.com which has been offering RHEL6 for a while but still has it >> > marked as "BETA" (it's not the RHEL6 beta, it's how they consider their >> > offering of RHEL6), which might indicate known problems between their >> > typical hardware and RHEL6. I didn't find any clues with a quick >> > bugzilla search... >> > >> are you using the bios raid or MD raid? ?If you are using biod raid you >> are running everything effectively pio mode and it's going to suck. ?Put >> the bios in ahci(not raid mode) and then use Linux raid from inside the >> partitioning scheme(which is md raid) i bet performance will be much better. > > I probably forgot to detail, but it's indeed MD RAID I'm using. No one > in their right mind would use one of those cheap motherboard pseudo > RAID, would they? :-) > > Anyway, I've reinstalled the server with RHEL5 and it's running fine > now. Definitely something going on between this particular hardware and > RHEL6.0... since I intend to use it only as a KVM hypervisor, and that > RHEL6 still doesn't support the new cool stuff that would make a > difference (like KSM), I'll stick with RHEL5, even though all of the > virtual servers running under it will be RHEL6. > > Matthias > > -- > Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ > Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 > Load : 0.07 0.07 0.08 > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From lowen at pari.edu Thu Feb 10 14:54:24 2011 From: lowen at pari.edu (Lamar Owen) Date: Thu, 10 Feb 2011 09:54:24 -0500 Subject: [rhelv6-list] Performance issues with RHEL6 (iowait) In-Reply-To: <20110209104721.0a971fee@python3.es.aed.lan> References: <20110209104721.0a971fee@python3.es.aed.lan> Message-ID: <201102100954.24523.lowen@pari.edu> On Wednesday, February 09, 2011 04:47:21 am Matthias Saou wrote: > it's a Supermicro X8STi motherboard (Xeon i7 920) with two cheap > SATA disks (WD10EADS-00P8B0) using software RAID-1. The WD EADS drives aren't usable in RAID configurations, by design, at least apparently with recent kernels. Google the string 'WDTLER' to find out more information. This issue is not isolated to Linux, either, with these low-end WD drives. WD has 'RE' drives that are set up in such a way to work well with RAID. I had one of the 1.5TB EADS drives in RAID 1 with a Seagate 1.5TB; the WD unit was constantly creating iowaits. I didn't try with the EL5 kernel; it's interesting that it works there and not with EL6. You could try the Fedora releases from FC6 up through F14 and see which kernel rev broke it; that would help troubleshoot this annoying issue. I would, but I no longer have that EADS drive available for testing. From tinnisg at gmail.com Thu Feb 10 15:13:39 2011 From: tinnisg at gmail.com (Tinnis G) Date: Thu, 10 Feb 2011 10:13:39 -0500 Subject: [rhelv6-list] Facing issue to register the server In-Reply-To: <4D508C12.4010606@lenhof.eu.org> References: <1297061491.2638.1.camel@vision.b2b4linux.nl> <4D4F9F0B.7090701@lenhof.eu.org> <6F56410FBED1FC41BCA804E16F594B0B019DDD@chvpkw8xmbx01.chvpk.chevrontexaco.net> <4D508C12.4010606@lenhof.eu.org> Message-ID: Hi , The problem is resolved now. It is silly mistake from my side. The entry NETMASK was missing in ifcfg-eth0. But the funny part is, everythng ( ssh, ftp, ping etc) are fine without this entry , only redat registration objected ( may be due to proxy ...) Thanks again to all... On Mon, Feb 7, 2011 at 7:19 PM, Jean-Yves LENHOF wrote: > Le 08/02/2011 00:42, Collins, Kevin [BEELINE] a ?crit : > > I assume you mean 'clock skew'? >> > > yep > my bad > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > -------------- next part -------------- An HTML attachment was scrubbed... URL: From christian.masopust at siemens.com Tue Feb 15 08:51:29 2011 From: christian.masopust at siemens.com (Masopust, Christian) Date: Tue, 15 Feb 2011 09:51:29 +0100 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... In-Reply-To: References: Message-ID: Short update to my kernel-panic.... it turns out that this panic is caused by a bug in RHEL6 kernel (all 2.6.32-71.*) and happens if somebody telnets to nlockmgr and simply press several times. newer kernels (in fedora 14) don't have this bug anymore. does probably anybody know a "bug number" or when this bug has been fixed in kernel? thanks a lot, christian p.s.: i'm already in contact with redhat-support, but don't have a fix up to now.... ________________________________ Von: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list-bounces at redhat.com] Im Auftrag von Masopust, Christian Gesendet: Mittwoch, 09. Februar 2011 12:01 An: 'Red Hat Enterprise Linux 6 (Santiago) discussion mailing-list' Betreff: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... Hi all, some of my RHEL6-systems are facing a kernel panic from time to time. 2 of them are huge HP's (DL585 G7) with 48cores and 128GB, one of them is an older Primergy RX300S2 (4 cores, 8GB). Some other systems (also Primergies) run fine all the time... some other facts: - all filesystems ext4 - nfs4 enabled - 3 bonding devices, each having 2 physical devices - 2 of the bonding devices configured for jumbo frames (MTU=9000) Here's the console-log from one of the HP's: ------------[ cut here ]------------ kernel BUG at fs/inode.c:1333! invalid opcode: 0000 [#1] SMP last sysfs file: /sys/devices/system/cpu/cpu47/cache/index2/shared_cpu_map CPU 4 Modules linked in: iptable_filter ip_tables nfs fscache fuse nfsd nfs_acl auth_rpcgss exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler lockd sunrpc bonding ipv6 dm_mirror dm_region_hash dm_log uinput power_meter hwmon bnx2 amd64_edac_mod edac_core edac_mce_amd i2c_piix4 sg h pilo nx_nic(U) ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic pata_atiixp ahci hpsa(U) radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core dm_mod [last unloaded: freq_table] Modules linked in: iptable_filter ip_tables nfs fscache fuse nfsd nfs_acl auth_rpcgss exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler lockd sunrpc bonding ipv6 dm_mirror dm_region_hash dm_log uinput power_meter hwmon bnx2 amd64_edac_mod edac_core edac_mce_amd i2c_piix4 sg h pilo nx_nic(U) ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic pata_atiixp ahci hpsa(U) radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core dm_mod [last unloaded: freq_table] Pid: 3393, comm: lockd Tainted: G W ---------------- 2.6.32-71.14.1.el6.x86_64 #1 ProLiant DL585 G7 RIP: 0010:[] [] iput+0x69/0x70 RSP: 0018:ffff88082b86fce0 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8802fc8616c8 RCX: 000000000000c60e RDX: ffff88202e13a901 RSI: ffffffffa0341de0 RDI: ffff8802fc8616c8 RBP: ffff88082b86fcf0 R08: 000000000002ac45 R09: 0000000000000000 R10: 000000000000000f R11: 0000000000000000 R12: ffff880227b49c00 R13: ffffffffa034e060 R14: ffff88202e13a940 R15: 00000000fffffff5 FS: 00007fac6a0247c0(0000) GS:ffff88002c240000(0000) knlGS:00000000f77916c0 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 00007fac6a048000 CR3: 0000000c2da36000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff4ff0 DR7: 0000000000000400 Process lockd (pid: 3393, threadinfo ffff88082b86e000, task ffff88082d9ab4e0) Stack: ffff88082b86fd40 ffff8802fc861680 ffff88082b86fd10 ffffffff813fdbf1 <0> ffff880227b49c00 ffff880227b49c00 ffff88082b86fd30 ffffffffa03351f8 <0> ffff88082b86fd30 ffff880227b49c10 ffff88082b86fd60 ffffffffa0341e2c Call Trace: [] sock_release+0x71/0x90 [] svc_sock_free+0x48/0x70 [sunrpc] [] svc_xprt_free+0x4c/0x70 [sunrpc] [] ? svc_xprt_free+0x0/0x70 [sunrpc] [] kref_put+0x37/0x70 [] svc_xprt_put+0x19/0x20 [sunrpc] [] svc_xprt_release+0xc1/0xe0 [sunrpc] [] svc_recv+0x2ed/0x830 [sunrpc] [] ? default_wake_function+0x0/0x20 [] lockd+0xc1/0x230 [lockd] [] ? lockd+0x0/0x230 [lockd] [] kthread+0x96/0xa0 [] child_rip+0xa/0x20 [] ? kthread+0x0/0xa0 [] ? child_rip+0x0/0x20 Code: 38 48 c7 c0 f0 7c 18 81 48 85 d2 74 12 48 8b 42 20 48 c7 c2 f0 7c 18 81 48 85 c0 48 0f 44 c2 48 89 df ff d0 48 83 c4 08 5b c9 c3 <0f> 0b eb fe 0f 1f 00 55 48 89 e5 41 55 41 54 53 48 83 ec 08 0f RIP [] iput+0x69/0x70 RSP ?Mounting proc filesystem Mounting sysfs filesystem Creating /dev Creating initial device nodes Free memory/Total memory (free %): 456164 / 495584 ( 92.0457 ) Loading jbd2.ko module Loading mbcache.ko module Loading ext4.ko module Loading crc-t10dif.ko module Loading sd_mod.ko module Loading ata_generic.ko module Loading exportfs.ko module Loading autofs4.ko module Loading ipmi_msghandler.ko module Loading sunrpc.ko module Loading ipv6.ko module Loading uinput.ko module Loading hwmon.ko module Loading bnx2.ko module Loading edac_core.ko module Loading edac_mce_amd.ko module Loading sg.ko module Loading hpilo.ko module Loading nx_nic.ko module Loading cdrom.ko module Loading pata_acpi.ko module Loading pata_atiixp.ko module Loading ahci.ko module Loading hpsa.ko module hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:03:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out hpsa 0000:44:00.0: controller message 03:00 timed out Loading i2c-core.ko module Loading dm-mod.ko module Loading nfs_acl.ko module Loading auth_rpcgss.ko module Loading ipmi_devintf.ko module Loading ipmi_si.ko module Loading lockd.ko module Loadingpower_meter ACPI000D:00: Ignoring unsafe software power cap! bonding.ko module Loading dm-log.ko module Loading power_meter.ko module Loading amd64_edac_mod.ko module Loading i2c-piix4.ko module Loading sr_mod.ko module Loading drm.ko module Loading i2c-algo-bit.ko module Loading nfsd.ko module Loading dm-region-hash.ko module Loading ttm.ko module Loading drm_kms_helper.ko module Loading dm-mirror.ko module Loading radeon.ko module Waiting for required block device discovery Waiting for 8 sdd-like device(s)...Found Creating Block Devices Creating block device loop0 Creating block device loop1 Creating block device loop2 Creating block device loop3 Creating block device loop4 Creating block device loop5 Creating block device loop6 Creating block device loop7 Creating block device ram0 Creating block device ram1 Creating block device ram10 Creating block device ram11 Creating block device ram12 Creating block device ram13 Creating block device ram14 Creating block device ram15 Creating block device ram2 Creating block device ram3 Creating block device ram4 Creating block device ram5 Creating block device ram6 Creating block device ram7 Creating block device ram8 Creating block device ram9 Creating block device sda Creating block device sdb Creating block device sdc Creating block device sdd Creating block device sr0 mdadm: No arrays found in config file or automatically Free memory/Total memory (free %): 432796 / 495584 ( 87.3305 ) Saving to the local filesystem /dev/sdd1 e2fsck 1.41.12 (17-May-2010) Homes: recovering journal Homes: clean, 9073003/164782080 files, 387571383/659105347 blocks Free memory/Total memory (free %): 427248 / 495584 ( 86.211 ) Copying data : [ 2 %] Copying data : [100 %] Saving core complete Restarting system. Backtrace from crash-dump utility shows: GNU gdb (GDB) 7.0 Copyright (C) 2009 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-unknown-linux-gnu"... KERNEL: /usr/lib/debug/lib/modules/2.6.32-71.14.1.el6.x86_64/vmlinux DUMPFILE: ./vmcore [PARTIAL DUMP] CPUS: 48 DATE: Wed Feb 9 09:30:52 2011 UPTIME: 14 days, 13:57:19 LOAD AVERAGE: 3.65, 3.39, 3.25 TASKS: 1663 NODENAME: hydra.sie.siemens.at RELEASE: 2.6.32-71.14.1.el6.x86_64 VERSION: #1 SMP Wed Jan 5 17:01:01 EST 2011 MACHINE: x86_64 (2095 Mhz) MEMORY: 128 GB PANIC: "kernel BUG at fs/inode.c:1333!" PID: 3393 COMMAND: "lockd" TASK: ffff88082d9ab4e0 [THREAD_INFO: ffff88082b86e000] CPU: 4 STATE: TASK_RUNNING (PANIC) crash> bt PID: 3393 TASK: ffff88082d9ab4e0 CPU: 4 COMMAND: "lockd" #0 [ffff88082b86f9a0] machine_kexec at ffffffff8103695b #1 [ffff88082b86fa00] crash_kexec at ffffffff810b9068 #2 [ffff88082b86fad0] oops_end at ffffffff814cc6e0 #3 [ffff88082b86fb00] die at ffffffff8101733b #4 [ffff88082b86fb30] do_trap at ffffffff814cbfb4 #5 [ffff88082b86fb90] do_invalid_op at ffffffff81014ee5 #6 [ffff88082b86fc30] invalid_op at ffffffff81013f5b [exception RIP: iput+105] RIP: ffffffff81186bf9 RSP: ffff88082b86fce0 RFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8802fc8616c8 RCX: 000000000000c60e RDX: ffff88202e13a901 RSI: ffffffffa0341de0 RDI: ffff8802fc8616c8 RBP: ffff88082b86fcf0 R8: 000000000002ac45 R9: 0000000000000000 R10: 000000000000000f R11: 0000000000000000 R12: ffff880227b49c00 R13: ffffffffa034e060 R14: ffff88202e13a940 R15: 00000000fffffff5 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 #7 [ffff88082b86fcf8] sock_release at ffffffff813fdbf1 #8 [ffff88082b86fd18] svc_sock_free at ffffffffa03351f8 #9 [ffff88082b86fd38] svc_xprt_free at ffffffffa0341e2c #10 [ffff88082b86fd68] kref_put at ffffffff8125cb97 #11 [ffff88082b86fd88] svc_xprt_put at ffffffffa0340f29 #12 [ffff88082b86fd98] svc_xprt_release at ffffffffa0341191 #13 [ffff88082b86fdc8] svc_recv at ffffffffa03415bd #14 [ffff88082b86fe58] lockd at ffffffffa02f6291 #15 [ffff88082b86fee8] kthread at ffffffff81091a76 #16 [ffff88082b86ff48] kernel_thread at ffffffff810141ca crash> any idea? any hint? what else can i do to find the reason for these panics? how to solve it? thanks a lot, christian -------------- next part -------------- An HTML attachment was scrubbed... URL: From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Tue Feb 15 09:39:43 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Tue, 15 Feb 2011 10:39:43 +0100 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... In-Reply-To: References: Message-ID: <20110215103943.3d96570b@python3.es.aed.lan> Masopust, Christian wrote : > Short update to my kernel-panic.... > > it turns out that this panic is caused by a bug in RHEL6 kernel (all 2.6.32-71.*) and happens > if somebody telnets to nlockmgr and simply press several times. > > newer kernels (in fedora 14) don't have this bug anymore. does probably anybody know a > "bug number" or when this bug has been fixed in kernel? Ouch! It used to be possible to find some testing EL kernels on various URLs under people.redhat.com, but it seems that some Red Hat internal policy has forbidden that from now on... that plus the new "self support" for the same price as having support previously, and you get a nice crippled entry level Enterprise OS. This makes me wonder : Is Red Hat intentionally driving its "lower-end" customers towards clones of it's own OS? It would seem logic to try and do the opposite, but hey, I'm no business expert... Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.05 0.05 0.05 From robinprice at gmail.com Tue Feb 15 23:37:57 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Tue, 15 Feb 2011 18:37:57 -0500 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... In-Reply-To: <20110215103943.3d96570b@python3.es.aed.lan> References: <20110215103943.3d96570b@python3.es.aed.lan> Message-ID: The test kernel urls are still there. Once I find the RHEL6 one, i'll post it here. RH is not driving any customers to RHEL clones. Here is RHEL5 for instance: http://people.redhat.com/jwilson/el5/ ~rp On Tue, Feb 15, 2011 at 4:39 AM, Matthias Saou wrote: > Masopust, Christian wrote : > >> Short update to my kernel-panic.... >> >> it turns out that this panic is caused by a bug in RHEL6 kernel (all 2.6.32-71.*) and happens >> if somebody telnets to nlockmgr and simply press several times. >> >> newer kernels (in fedora 14) don't have this bug anymore. ?does probably anybody know a >> "bug number" or when this bug has been fixed in kernel? > > Ouch! It used to be possible to find some testing EL kernels on various > URLs under people.redhat.com, but it seems that some Red Hat internal > policy has forbidden that from now on... that plus the new "self > support" for the same price as having support previously, and you get a > nice crippled entry level Enterprise OS. > > This makes me wonder : Is Red Hat intentionally driving its "lower-end" > customers towards clones of it's own OS? It would seem logic to try and > do the opposite, but hey, I'm no business expert... > > Matthias > > -- > Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ > Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 > Load : 0.05 0.05 0.05 > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 16 10:55:46 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 16 Feb 2011 11:55:46 +0100 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... In-Reply-To: References: <20110215103943.3d96570b@python3.es.aed.lan> Message-ID: <20110216115546.45f20960@python3.es.aed.lan> robinprice at gmail.com wrote : > The test kernel urls are still there. Once I find the RHEL6 one, i'll > post it here. RH is not driving any customers to RHEL clones. > > Here is RHEL5 for instance: > http://people.redhat.com/jwilson/el5/ Yes, RHEL5 kernels are still here and there... RHEL6 aren't. You'll just find pre-6.0-final kernels on people.redhat.com. See for instance : http://people.redhat.com/linville/kernels/rhel6/ "Due to changes in Red Hat policy regarding release of test kernels, I am no longer at liberty to release test kernels to the general public. If this is inconvenient for you, please register your complaint through whatever Red Hat support means are available to you." This s*cks and I would encourage everyone to do what John Linville suggests in order to try and get this policy reverted. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.05 0.09 0.11 From robinprice at gmail.com Wed Feb 16 12:43:43 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Wed, 16 Feb 2011 07:43:43 -0500 Subject: [rhelv6-list] RHEL6 kernel 2.6.32-71.14.1.el6.x86_64 panic... In-Reply-To: <20110216115546.45f20960@python3.es.aed.lan> References: <20110215103943.3d96570b@python3.es.aed.lan> <20110216115546.45f20960@python3.es.aed.lan> Message-ID: Good to know. I will try dig around and see what I can find. Thanks. ~rp On Wed, Feb 16, 2011 at 5:55 AM, Matthias Saou wrote: > robinprice at gmail.com wrote : > >> The test kernel urls are still there. ?Once I find the RHEL6 one, i'll >> post it here. ?RH is not driving any customers to RHEL clones. >> >> Here is RHEL5 for instance: >> http://people.redhat.com/jwilson/el5/ > > Yes, RHEL5 kernels are still here and there... RHEL6 aren't. You'll > just find pre-6.0-final kernels on people.redhat.com. See for instance : > > http://people.redhat.com/linville/kernels/rhel6/ > > "Due to changes in Red Hat policy regarding release of test kernels, I > am no longer at liberty to release test kernels to the general public. > > If this is inconvenient for you, please register your complaint through > whatever Red Hat support means are available to you." > > This s*cks and I would encourage everyone to do what John Linville > suggests in order to try and get this policy reverted. > > Matthias > > -- > Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ > Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 > Load : 0.05 0.09 0.11 > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From carlopmart at gmail.com Fri Feb 18 09:36:00 2011 From: carlopmart at gmail.com (carlopmart) Date: Fri, 18 Feb 2011 10:36:00 +0100 Subject: [rhelv6-list] ifcfg-ipsec interfaces dosen't works on rhel6 Message-ID: <4D5E3D80.3090306@gmail.com> Hi all, Does Redhat disabled ifcfg-ipsecX files configuration to establish ipsec tunnels on RHEL6?? Under RHEL5 all works very well ... Do I need to do manual configuration using openswan package?? Many thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From carlopmart at gmail.com Mon Feb 21 09:50:49 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 10:50:49 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 Message-ID: <4D623579.70605@gmail.com> Hi all, I need to resize tmpfs filesystem to use 66% of memory but i can't. I have changed fstab to include "size=66%" in tmpfs configuration (I have tried to use "size=2G" also with the same result) and nothing works. My laptop has 3GB of RAM. Is this a bug or am I doing some mistake?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From christian.masopust at siemens.com Mon Feb 21 11:46:01 2011 From: christian.masopust at siemens.com (Masopust, Christian) Date: Mon, 21 Feb 2011 12:46:01 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D623579.70605@gmail.com> References: <4D623579.70605@gmail.com> Message-ID: dumb question.... did you remount tmpfs after change in fstab? christian > -----Urspr?ngliche Nachricht----- > Von: rhelv6-list-bounces at redhat.com > [mailto:rhelv6-list-bounces at redhat.com] Im Auftrag von carlopmart > Gesendet: Montag, 21. Februar 2011 10:51 > An: rhelv6-list at redhat.com > Betreff: [rhelv6-list] Problems resizing tmpfs under RHEL6 > > Hi all, > > I need to resize tmpfs filesystem to use 66% of memory but > i can't. I have changed > fstab to include "size=66%" in tmpfs configuration (I have > tried to use "size=2G" > also with the same result) and nothing works. My laptop has > 3GB of RAM. Is this a > bug or am I doing some mistake?? > > Thanks. > > -- > CL Martinez > carlopmart {at} gmail {d0t} com > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From carlopmart at gmail.com Mon Feb 21 12:17:39 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 13:17:39 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: References: <4D623579.70605@gmail.com> Message-ID: <4D6257E3.8060805@gmail.com> On 02/21/2011 12:46 PM, Masopust, Christian wrote: > > dumb question.... did you remount tmpfs after change in fstab? > > christian > Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of memory ... -- CL Martinez carlopmart {at} gmail {d0t} com From notting at redhat.com Mon Feb 21 17:02:26 2011 From: notting at redhat.com (Bill Nottingham) Date: Mon, 21 Feb 2011 12:02:26 -0500 Subject: [rhelv6-list] ifcfg-ipsec interfaces dosen't works on rhel6 In-Reply-To: <4D5E3D80.3090306@gmail.com> References: <4D5E3D80.3090306@gmail.com> Message-ID: <20110221170225.GA8811@nostromo.devel.redhat.com> carlopmart (carlopmart at gmail.com) said: > Does Redhat disabled ifcfg-ipsecX files configuration to establish > ipsec tunnels on RHEL6?? Under RHEL5 all works very well ... Do I > need to do manual configuration using openswan package?? Yes, the IPSec infrastructure switched to using openswan, which uses its own configuration format. Bill From carlopmart at gmail.com Mon Feb 21 17:32:21 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 18:32:21 +0100 Subject: [rhelv6-list] ifcfg-ipsec interfaces dosen't works on rhel6 In-Reply-To: <20110221170225.GA8811@nostromo.devel.redhat.com> References: <4D5E3D80.3090306@gmail.com> <20110221170225.GA8811@nostromo.devel.redhat.com> Message-ID: <4D62A1A5.1050807@gmail.com> On 02/21/2011 06:02 PM, Bill Nottingham wrote: > carlopmart (carlopmart at gmail.com) said: >> Does Redhat disabled ifcfg-ipsecX files configuration to establish >> ipsec tunnels on RHEL6?? Under RHEL5 all works very well ... Do I >> need to do manual configuration using openswan package?? > > Yes, the IPSec infrastructure switched to using openswan, > which uses its own configuration format. > > Bill > Thanks Bill. -- CL Martinez carlopmart {at} gmail {d0t} com From carlopmart at gmail.com Mon Feb 21 18:56:35 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 19:56:35 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D6257E3.8060805@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> Message-ID: <4D62B563.8080806@gmail.com> On 02/21/2011 01:17 PM, carlopmart wrote: > On 02/21/2011 12:46 PM, Masopust, Christian wrote: >> >> dumb question.... did you remount tmpfs after change in fstab? >> >> christian >> > > Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of memory ... > Please, any ideas?? -- CL Martinez carlopmart {at} gmail {d0t} com From brilong at cisco.com Mon Feb 21 20:21:05 2011 From: brilong at cisco.com (Brian Long) Date: Mon, 21 Feb 2011 15:21:05 -0500 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D62B563.8080806@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> Message-ID: <4D62C931.9070309@cisco.com> On 2/21/11 1:56 PM, carlopmart wrote: > On 02/21/2011 01:17 PM, carlopmart wrote: >> On 02/21/2011 12:46 PM, Masopust, Christian wrote: >>> >>> dumb question.... did you remount tmpfs after change in fstab? >>> >>> christian >>> >> >> Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of >> memory ... >> > > Please, any ideas?? The mount man page says the following: "The tmpfs mount options for sizing ( size, nr_blocks, and nr_inodes) accept a suffix k, m or g for Ki, Mi, Gi (binary kilo, mega and giga) and can be changed on remount." If % does not work in RHEL 6, have you tried "2g" instead of "2G"? :-) /Brian/ -- Brian Long | | Corporate Security Programs Org . | | | . | | | . ' ' C I S C O From jfranz at freerun.com Mon Feb 21 22:10:41 2011 From: jfranz at freerun.com (Benjamin Franz) Date: Mon, 21 Feb 2011 14:10:41 -0800 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D62C931.9070309@cisco.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> Message-ID: <4D62E2E1.2030205@freerun.com> On 02/21/2011 12:21 PM, Brian Long wrote: > On 2/21/11 1:56 PM, carlopmart wrote: >> On 02/21/2011 01:17 PM, carlopmart wrote: >>> Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of >>> memory ... >> Please, any ideas?? > The mount man page says the following: > "The tmpfs mount options for sizing ( size, nr_blocks, and nr_inodes) > accept a suffix k, m or g for Ki, Mi, Gi (binary kilo, mega and > giga) and can be changed on remount." > > If % does not work in RHEL 6, have you tried "2g" instead of "2G"? :-) +1 I have /dev/shm set to 8g on one of my machines using 'size=8g'. -- Benjamin Franz From carlopmart at gmail.com Mon Feb 21 22:22:41 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 23:22:41 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D62C931.9070309@cisco.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> Message-ID: <4D62E5B1.7060703@gmail.com> On 02/21/2011 09:21 PM, Brian Long wrote: > On 2/21/11 1:56 PM, carlopmart wrote: >> On 02/21/2011 01:17 PM, carlopmart wrote: >>> On 02/21/2011 12:46 PM, Masopust, Christian wrote: >>>> >>>> dumb question.... did you remount tmpfs after change in fstab? >>>> >>>> christian >>>> >>> >>> Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of >>> memory ... >>> >> >> Please, any ideas?? > > The mount man page says the following: > "The tmpfs mount options for sizing ( size, nr_blocks, and nr_inodes) > accept a suffix k, m or g for Ki, Mi, Gi (binary kilo, mega and > giga) and can be changed on remount." > > If % does not work in RHEL 6, have you tried "2g" instead of "2G"? :-) > > /Brian/ Yes, I have tried all options like in man pages is explained ... without luck. My fstab: UUID=ea53b08b-1557-4e37-b88a-e3b672bcf569 / ext4 defaults 1 1 /dev/mapper/cryptvol-datavol /export/data ext4 defaults 1 2 /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 /export/home ext4 defaults 1 2 UUID=d16b644b-020f-4693-bc21-e5d9aa9b8e7b swap swap defaults 0 0 tmpfs /dev/shm tmpfs size=2g 0 0 devpts /dev/pts devpts gid=5,mode=620 0 0 sysfs /sys sysfs defaults 0 0 proc /proc proc defaults 0 0 -- CL Martinez carlopmart {at} gmail {d0t} com From carlopmart at gmail.com Mon Feb 21 22:23:38 2011 From: carlopmart at gmail.com (carlopmart) Date: Mon, 21 Feb 2011 23:23:38 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D62E2E1.2030205@freerun.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> Message-ID: <4D62E5EA.4060802@gmail.com> On 02/21/2011 11:10 PM, Benjamin Franz wrote: > On 02/21/2011 12:21 PM, Brian Long wrote: >> On 2/21/11 1:56 PM, carlopmart wrote: >>> On 02/21/2011 01:17 PM, carlopmart wrote: >>>> Yes, and works. But if I reboot my laptop, tmpfs uses only 50% of memory ... >>> Please, any ideas?? >> The mount man page says the following: >> "The tmpfs mount options for sizing ( size, nr_blocks, and nr_inodes) >> accept a suffix k, m or g for Ki, Mi, Gi (binary kilo, mega and >> giga) and can be changed on remount." >> >> If % does not work in RHEL 6, have you tried "2g" instead of "2G"? :-) > > +1 > > I have /dev/shm set to 8g on one of my machines using 'size=8g'. > Can you show your fstab entry for tmpfs Bejamin?? Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From jfranz at freerun.com Tue Feb 22 13:44:14 2011 From: jfranz at freerun.com (Benjamin Franz) Date: Tue, 22 Feb 2011 05:44:14 -0800 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D62E5EA.4060802@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> Message-ID: <4D63BDAE.3080900@freerun.com> On 02/21/2011 02:23 PM, carlopmart wrote: > On 02/21/2011 11:10 PM, Benjamin Franz wrote: >> >> I have /dev/shm set to 8g on one of my machines using 'size=8g'. >> > > Can you show your fstab entry for tmpfs Bejamin?? tmpfs /dev/shm tmpfs size=8g 0 0 Note: This is on a CentOS5.5 machine. I haven't tried it on a RH6 machine. -- Benjamin Franz From carlopmart at gmail.com Tue Feb 22 14:29:48 2011 From: carlopmart at gmail.com (carlopmart) Date: Tue, 22 Feb 2011 15:29:48 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63BDAE.3080900@freerun.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> Message-ID: <4D63C85C.70308@gmail.com> On 02/22/2011 02:44 PM, Benjamin Franz wrote: > On 02/21/2011 02:23 PM, carlopmart wrote: >> On 02/21/2011 11:10 PM, Benjamin Franz wrote: >>> >>> I have /dev/shm set to 8g on one of my machines using 'size=8g'. >>> >> >> Can you show your fstab entry for tmpfs Bejamin?? > > tmpfs /dev/shm tmpfs size=8g 0 0 > > Note: This is on a CentOS5.5 machine. I haven't tried it on a RH6 machine. > Ok, and yes. Under centos 5 or rhel5 all works as expected (i have another rhel5 host and tmpfs works ok), but under rhel6 not. For example. My fstab: # # /etc/fstab # Created by anaconda on Sun Feb 20 14:33:14 2011 # # Accessible filesystems, by reference, are maintained under '/dev/disk' # See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info # UUID=ea53b08b-1557-4e37-b88a-e3b672bcf569 / ext4 defaults 1 1 /dev/mapper/cryptvol-datavol /export/data ext4 defaults 1 2 /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 /export/home ext4 defaults 1 2 UUID=d16b644b-020f-4693-bc21-e5d9aa9b8e7b swap swap defaults 0 0 tmpfs /dev/shm tmpfs size=2g 0 0 devpts /dev/pts devpts gid=5,mode=620 0 0 sysfs /sys sysfs defaults 0 0 proc /proc proc defaults 0 0 Output of mount command: [carlos at rhel6srv tmp]$ mount /dev/sda1 on / type ext4 (rw) proc on /proc type proc (rw) sysfs on /sys type sysfs (rw) devpts on /dev/pts type devpts (rw,gid=5,mode=620) tmpfs on /dev/shm type tmpfs (rw,size=2g) /dev/mapper/cryptvol-datavol on /export/data type ext4 (rw) /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 on /export/home type ext4 (rw) none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw) So far, everything is fine. But when I execute df command: [carlos at rhel6srv tmp]$ df -h Filesystem Size Used Avail Use% Mounted on /dev/sda1 7.9G 2.7G 4.9G 36% / tmpfs 1.5G 312K 1.5G 1% /dev/shm /dev/mapper/cryptvol-datavol 63G 9.6G 51G 17% /export/data /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 16G 3.7G 12G 25% /export/home .. shows only 50% of memory for tmpfs. Weird, right?? -- CL Martinez carlopmart {at} gmail {d0t} com From jfranz at freerun.com Tue Feb 22 14:58:40 2011 From: jfranz at freerun.com (Jerry Franz) Date: Tue, 22 Feb 2011 06:58:40 -0800 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63C85C.70308@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> Message-ID: <4D63CF20.8090007@freerun.com> On 02/22/2011 06:29 AM, carlopmart wrote: > > [carlos at rhel6srv tmp]$ df -h > Filesystem Size Used Avail Use% Mounted on > /dev/sda1 7.9G 2.7G 4.9G 36% / > tmpfs 1.5G 312K 1.5G 1% /dev/shm > /dev/mapper/cryptvol-datavol > 63G 9.6G 51G 17% /export/data > /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 > 16G 3.7G 12G 25% /export/home > > .. shows only 50% of memory for tmpfs. Weird, right?? > Check /etc/sysctl.conf. You may need to add/change the kernel.shmmax and/or kernel.shmall values. -- Benjamin Franz From carlopmart at gmail.com Tue Feb 22 15:38:06 2011 From: carlopmart at gmail.com (carlopmart) Date: Tue, 22 Feb 2011 16:38:06 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63CF20.8090007@freerun.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <4D63CF20.8090007@freerun.com> Message-ID: <4D63D85E.10501@gmail.com> On 02/22/2011 03:58 PM, Jerry Franz wrote: > On 02/22/2011 06:29 AM, carlopmart wrote: >> >> [carlos at rhel6srv tmp]$ df -h >> Filesystem Size Used Avail Use% Mounted on >> /dev/sda1 7.9G 2.7G 4.9G 36% / >> tmpfs 1.5G 312K 1.5G 1% /dev/shm >> /dev/mapper/cryptvol-datavol >> 63G 9.6G 51G 17% /export/data >> /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 >> 16G 3.7G 12G 25% /export/home >> >> .. shows only 50% of memory for tmpfs. Weird, right?? >> > > > Check /etc/sysctl.conf. You may need to add/change the kernel.shmmax and/or > kernel.shmall values. > Uhmm why?? With rhel5 isn't needed ... But, what values do I need to put?? Actually: kernel.shmmax = 33554432 kernel.shmall = 2097152 -- CL Martinez carlopmart {at} gmail {d0t} com From rhelv6-list at open-pla.net Tue Feb 22 15:44:06 2011 From: rhelv6-list at open-pla.net (Udo) Date: Tue, 22 Feb 2011 16:44:06 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63C85C.70308@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> Message-ID: <1298389446.12189.8.camel@localhost.localdomain> Hello, my RHEL6 works as expected (RAM 3GB, size=50% is the default): tmpfs /dev/shm tmpfs defaults 0 0 tmpfs 1,5G 26M 1,5G 2% /dev/shm tmpfs /dev/shm tmpfs defaults,size=66% 0 0 tmpfs 2,0G 26M 1,9G 2% /dev/shm tmpfs /dev/shm tmpfs defaults,size=33% 0 0 tmpfs 984M 26M 959M 3% /dev/shm Also with the size=Xg notation. Udo Am Dienstag, den 22.02.2011, 15:29 +0100 schrieb carlopmart: > On 02/22/2011 02:44 PM, Benjamin Franz wrote: > > On 02/21/2011 02:23 PM, carlopmart wrote: > >> On 02/21/2011 11:10 PM, Benjamin Franz wrote: > >>> > >>> I have /dev/shm set to 8g on one of my machines using 'size=8g'. > >>> > >> > >> Can you show your fstab entry for tmpfs Bejamin?? > > > > tmpfs /dev/shm tmpfs size=8g 0 0 > > > > Note: This is on a CentOS5.5 machine. I haven't tried it on a RH6 machine. > > > > Ok, and yes. Under centos 5 or rhel5 all works as expected (i have another rhel5 > host and tmpfs works ok), but under rhel6 not. For example. > > My fstab: > > # > # /etc/fstab > # Created by anaconda on Sun Feb 20 14:33:14 2011 > # > # Accessible filesystems, by reference, are maintained under '/dev/disk' > # See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info > # > UUID=ea53b08b-1557-4e37-b88a-e3b672bcf569 / ext4 defaults > 1 1 > /dev/mapper/cryptvol-datavol /export/data ext4 defaults 1 2 > /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 /export/home ext4 > defaults 1 2 > UUID=d16b644b-020f-4693-bc21-e5d9aa9b8e7b swap swap defaults > 0 0 > tmpfs /dev/shm tmpfs size=2g 0 0 > devpts /dev/pts devpts gid=5,mode=620 0 0 > sysfs /sys sysfs defaults 0 0 > proc /proc proc defaults 0 0 > > Output of mount command: > > [carlos at rhel6srv tmp]$ mount > /dev/sda1 on / type ext4 (rw) > proc on /proc type proc (rw) > sysfs on /sys type sysfs (rw) > devpts on /dev/pts type devpts (rw,gid=5,mode=620) > tmpfs on /dev/shm type tmpfs (rw,size=2g) > /dev/mapper/cryptvol-datavol on /export/data type ext4 (rw) > /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 on /export/home type ext4 (rw) > none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw) > > So far, everything is fine. But when I execute df command: > > [carlos at rhel6srv tmp]$ df -h > Filesystem Size Used Avail Use% Mounted on > /dev/sda1 7.9G 2.7G 4.9G 36% / > tmpfs 1.5G 312K 1.5G 1% /dev/shm > /dev/mapper/cryptvol-datavol > 63G 9.6G 51G 17% /export/data > /dev/mapper/luks-ac5ed44c-f316-4b3a-9b2d-c3c15228ce04 > 16G 3.7G 12G 25% /export/home > > .. shows only 50% of memory for tmpfs. Weird, right?? > > From prentice at ias.edu Tue Feb 22 15:56:01 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Tue, 22 Feb 2011 10:56:01 -0500 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63D85E.10501@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <4D63CF20.8090007@freerun.com> <4D63D85E.10501@gmail.com> Message-ID: <4D63DC91.7070808@ias.edu> carlopmart wrote: > On 02/22/2011 03:58 PM, Jerry Franz wrote: >> On 02/22/2011 06:29 AM, carlopmart wrote: > > Uhmm why?? With rhel5 isn't needed ... But this isn't RHEL5, is it? -- Prentice From carlopmart at gmail.com Tue Feb 22 17:13:15 2011 From: carlopmart at gmail.com (carlopmart) Date: Tue, 22 Feb 2011 18:13:15 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63DC91.7070808@ias.edu> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <4D63CF20.8090007@freerun.com> <4D63D85E.10501@gmail.com> <4D63DC91.7070808@ias.edu> Message-ID: <4D63EEAB.6090900@gmail.com> On 02/22/2011 04:56 PM, Prentice Bisbal wrote: > carlopmart wrote: >> On 02/22/2011 03:58 PM, Jerry Franz wrote: >>> On 02/22/2011 06:29 AM, carlopmart wrote: >> >> Uhmm why?? With rhel5 isn't needed ... > > But this isn't RHEL5, is it? > No, it is a RHEL6 ... -- CL Martinez carlopmart {at} gmail {d0t} com From carlopmart at gmail.com Tue Feb 22 17:15:21 2011 From: carlopmart at gmail.com (carlopmart) Date: Tue, 22 Feb 2011 18:15:21 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <1298389446.12189.8.camel@localhost.localdomain> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> Message-ID: <4D63EF29.5010707@gmail.com> On 02/22/2011 04:44 PM, Udo wrote: > Hello, > > my RHEL6 works as expected (RAM 3GB, size=50% is the default): > > tmpfs /dev/shm tmpfs defaults 0 0 > tmpfs 1,5G 26M 1,5G 2% /dev/shm > > tmpfs /dev/shm tmpfs defaults,size=66% 0 0 > tmpfs 2,0G 26M 1,9G 2% /dev/shm > > tmpfs /dev/shm tmpfs defaults,size=33% 0 0 > tmpfs 984M 26M 959M 3% /dev/shm > > > Also with the size=Xg notation. > > > Udo > > Incredible. I don't understand ... And, do you have modified something apart of fstab, Udo?? Is your rhel6 i386 or x86_64?? Mine is i386. -- CL Martinez carlopmart {at} gmail {d0t} com From rhelv6-list at open-pla.net Wed Feb 23 08:34:17 2011 From: rhelv6-list at open-pla.net (Udo) Date: Wed, 23 Feb 2011 09:34:17 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D63EF29.5010707@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> Message-ID: <1298450057.12034.4.camel@localhost.localdomain> Am Dienstag, den 22.02.2011, 18:15 +0100 schrieb carlopmart: > On 02/22/2011 04:44 PM, Udo wrote: > > Hello, > > > > my RHEL6 works as expected (RAM 3GB, size=50% is the default): > > > > tmpfs /dev/shm tmpfs defaults 0 0 > > tmpfs 1,5G 26M 1,5G 2% /dev/shm > > > > tmpfs /dev/shm tmpfs defaults,size=66% 0 0 > > tmpfs 2,0G 26M 1,9G 2% /dev/shm > > > > tmpfs /dev/shm tmpfs defaults,size=33% 0 0 > > tmpfs 984M 26M 959M 3% /dev/shm > > > > > > Also with the size=Xg notation. > > > > > > Udo > > > > > Incredible. I don't understand ... And, do you have modified something apart of > fstab, Udo?? Is your rhel6 i386 or x86_64?? Mine is i386. > > Hi, it is a freshly installed RHEL6-x86_64 from iso without updates. The changes are only in /etc/fstab, as you can see above, and a remount command. Udo -------------- next part -------------- An HTML attachment was scrubbed... URL: From carlopmart at gmail.com Wed Feb 23 09:13:06 2011 From: carlopmart at gmail.com (carlopmart) Date: Wed, 23 Feb 2011 10:13:06 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <1298450057.12034.4.camel@localhost.localdomain> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> <1298450057.12034.4.camel@localhost.localdomain> Message-ID: <4D64CFA2.4070103@gmail.com> On 02/23/2011 09:34 AM, Udo wrote: > > Am Dienstag, den 22.02.2011, 18:15 +0100 schrieb carlopmart: >> On 02/22/2011 04:44 PM, Udo wrote: >> > Hello, >> > >> > my RHEL6 works as expected (RAM 3GB, size=50% is the default): >> > >> > tmpfs /dev/shm tmpfs defaults 0 0 >> > tmpfs 1,5G 26M 1,5G 2% /dev/shm >> > >> > tmpfs /dev/shm tmpfs defaults,size=66% 0 0 >> > tmpfs 2,0G 26M 1,9G 2% /dev/shm >> > >> > tmpfs /dev/shm tmpfs defaults,size=33% 0 0 >> > tmpfs 984M 26M 959M 3% /dev/shm >> > >> > >> > Also with the size=Xg notation. >> > >> > >> > Udo >> > >> > >> Incredible. I don't understand ... And, do you have modified something apart of >> fstab, Udo?? Is your rhel6 i386 or x86_64?? Mine is i386. >> >> > > Hi, > > it is a freshly installed RHEL6-x86_64 from iso without updates. > The changes are only in /etc/fstab, as you can see above, and a remount command. > > Udo > > Do you need to execute remount command after host is up?? I need to run every time when laptop is up. Thanks. -- CL Martinez carlopmart {at} gmail {d0t} com From rhelv6-list at open-pla.net Wed Feb 23 10:28:45 2011 From: rhelv6-list at open-pla.net (Udo) Date: Wed, 23 Feb 2011 11:28:45 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D64CFA2.4070103@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> <1298450057.12034.4.camel@localhost.localdomain> <4D64CFA2.4070103@gmail.com> Message-ID: <1298456925.12034.18.camel@localhost.localdomain> Am Mittwoch, den 23.02.2011, 10:13 +0100 schrieb carlopmart: > > > > Hi, > > > > it is a freshly installed RHEL6-x86_64 from iso without updates. > > The changes are only in /etc/fstab, as you can see above, and a remount command. > > > > Udo > > > > > > Do you need to execute remount command after host is up?? I need to run every time > when laptop is up. > > Thanks. > Oh sorry, I didn't realized that you only have the problem after a reboot. After a reboot I have the same effect. But the following patch correct it. The backshlashes marks the linebreak. --- /etc/rc.d/rc.sysinit.orig 2010-09-01 18:15:50.000000000 +0200 +++ /etc/rc.d/rc.sysinit 2011-02-23 11:07:12.544475820 +0100 @@ -487,7 +487,7 @@ mount -f /proc >/dev/null 2>&1 mount -f /sys >/dev/null 2>&1 mount -f /dev/pts >/dev/null 2>&1 - mount -f /dev/shm >/dev/null 2>&1 + #mount -f /dev/shm >/dev/null 2>&1 mount -f /proc/bus/usb >/dev/null 2>&1 fi @@ -495,7 +495,7 @@ # mounted). Contrary to standard usage, # filesystems are NOT unmounted in single user mode. if [ "$READONLY" != "yes" ] ; then - action $"Mounting local filesystems: " mount -a -t \ nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev + action $"Mounting local filesystems: " mount -a -t \ tmpfs,nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev else action $"Mounting local filesystems: " mount -a -n -t \ nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev fi If the mount point is different, it works as expected after a reboot (unpatched rc.sysinit). cat /etc/fstab |grep tmpfs tmpfs /dev/shm tmpfs defaults,size=66% 0 0 tmpfs /mnt/tmpfs tmpfs defaults,size=33% 0 0 df -h | grep tmpfs tmpfs 1,5G 26M 1,5G 2% /dev/shm tmpfs 984M 0 984M 0% /mnt/tmpfs It's a bug or a feature? The command /bin/mount -o remount /dev/shm in /etc/rc.d/rc.local solve your problem. Udo From carlopmart at gmail.com Wed Feb 23 10:47:46 2011 From: carlopmart at gmail.com (carlopmart) Date: Wed, 23 Feb 2011 11:47:46 +0100 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <1298456925.12034.18.camel@localhost.localdomain> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> <1298450057.12034.4.camel@localhost.localdomain> <4D64CFA2.4070103@gmail.com> <1298456925.12034.18.camel@localhost.localdomain> Message-ID: <4D64E5D2.20208@gmail.com> On 02/23/2011 11:28 AM, Udo wrote: > Am Mittwoch, den 23.02.2011, 10:13 +0100 schrieb carlopmart: >>> >>> Hi, >>> >>> it is a freshly installed RHEL6-x86_64 from iso without updates. >>> The changes are only in /etc/fstab, as you can see above, and a remount command. >>> >>> Udo >>> >>> >> >> Do you need to execute remount command after host is up?? I need to run every time >> when laptop is up. >> >> Thanks. >> > > Oh sorry, > > I didn't realized that you only have the problem after a reboot. > After a reboot I have the same effect. > > But the following patch correct it. > The backshlashes marks the linebreak. > > --- /etc/rc.d/rc.sysinit.orig 2010-09-01 18:15:50.000000000 +0200 > +++ /etc/rc.d/rc.sysinit 2011-02-23 11:07:12.544475820 +0100 > @@ -487,7 +487,7 @@ > mount -f /proc>/dev/null 2>&1 > mount -f /sys>/dev/null 2>&1 > mount -f /dev/pts>/dev/null 2>&1 > - mount -f /dev/shm>/dev/null 2>&1 > + #mount -f /dev/shm>/dev/null 2>&1 > mount -f /proc/bus/usb>/dev/null 2>&1 > fi > > @@ -495,7 +495,7 @@ > # mounted). Contrary to standard usage, > # filesystems are NOT unmounted in single user mode. > if [ "$READONLY" != "yes" ] ; then > - action $"Mounting local filesystems: " mount -a -t \ > nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev > + action $"Mounting local filesystems: " mount -a -t \ > tmpfs,nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev > else > action $"Mounting local filesystems: " mount -a -n -t \ > nonfs,nfs4,smbfs,ncpfs,cifs,gfs,gfs2 -O no_netdev > fi > > > If the mount point is different, it works as expected after a reboot > (unpatched rc.sysinit). > > cat /etc/fstab |grep tmpfs > tmpfs /dev/shm tmpfs defaults,size=66% 0 0 > tmpfs /mnt/tmpfs tmpfs defaults,size=33% 0 0 > > df -h | grep tmpfs > tmpfs 1,5G 26M 1,5G 2% /dev/shm > tmpfs 984M 0 984M 0% /mnt/tmpfs > > > It's a bug or a feature? > > > The command /bin/mount -o remount /dev/shm > in /etc/rc.d/rc.local solve your problem. > > > Udo > > In my opinion it is a bug, because under RHEL5.x works perfectly. Many thanks Udo. -- CL Martinez carlopmart {at} gmail {d0t} com From john.haxby at gmail.com Wed Feb 23 12:56:46 2011 From: john.haxby at gmail.com (John Haxby) Date: Wed, 23 Feb 2011 12:56:46 +0000 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D64E5D2.20208@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> <1298450057.12034.4.camel@localhost.localdomain> <4D64CFA2.4070103@gmail.com> <1298456925.12034.18.camel@localhost.localdomain> <4D64E5D2.20208@gmail.com> Message-ID: On 23 February 2011 10:47, carlopmart wrote: > > In my opinion it is a bug, because under RHEL5.x works perfectly. > > /dev/shm is mounted in the initramfs (/boot/initramfs...). The bug is in dracut -- either it should copy /etc/fstab into the initramfs (easy) or it should extract the mount options (harder). Either way to make it stick you can run "dracut -f" to make the change stick after a reboot. Although I haven't tried this, you can probably make it work by patching up your initamfs. The file /boot/initramfs-`uname -r`.img is a gzipped cpio archive so you should be able to unwrap it, pop /etc/fstab in its etc directory and wrap it up again. Of course, you need to be careful because it's easy to wind up with an unbootable system if (or when) you get it wrong. jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From jcm at redhat.com Wed Feb 23 16:25:04 2011 From: jcm at redhat.com (Jon Masters) Date: Wed, 23 Feb 2011 11:25:04 -0500 Subject: [rhelv6-list] Problems resizing tmpfs under RHEL6 In-Reply-To: <4D64E5D2.20208@gmail.com> References: <4D623579.70605@gmail.com> <4D6257E3.8060805@gmail.com> <4D62B563.8080806@gmail.com> <4D62C931.9070309@cisco.com> <4D62E2E1.2030205@freerun.com> <4D62E5EA.4060802@gmail.com> <4D63BDAE.3080900@freerun.com> <4D63C85C.70308@gmail.com> <1298389446.12189.8.camel@localhost.localdomain> <4D63EF29.5010707@gmail.com> <1298450057.12034.4.camel@localhost.localdomain> <4D64CFA2.4070103@gmail.com> <1298456925.12034.18.camel@localhost.localdomain> <4D64E5D2.20208@gmail.com> Message-ID: <1298478304.15514.20.camel@constitution.bos.jonmasters.org> On Wed, 2011-02-23 at 11:47 +0100, carlopmart wrote: > In my opinion it is a bug, because under RHEL5.x works perfectly. Please open a bug against RHEL6 for this. Jon. From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Feb 23 16:59:10 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 23 Feb 2011 17:59:10 +0100 Subject: [rhelv6-list] Subversion client with https and internal CA on RHEL6 Message-ID: <20110223175910.053e059d@python3.es.aed.lan> Hi, I can't seem to figure out the proper clean way to have the svn CLI client trust all https URLs using certificates signed by an internal CA. With RHEL5, it was easy : $ strace svn co https://myserver/repo/ 2>&1 | grep pki open("/etc/pki/tls/cert.pem", O_RDONLY) = 3 open("/etc/pki/tls/cert.pem", O_RDONLY) = 3 stat("/etc/pki/tls/certs/b903d65c.0", 0x7fff7f839980) = -1 ENOENT (No such file or directory) The SSL library being used looked for a CA certificate named after a hash specific to that certificate (which you got with openssl x509 -hash -noout -in myca.crt). With RHEL6 this happens no more : $ strace svn co https://myserver/repo/ 2>&1 | grep pki open("/etc/pki/tls/certs/ca-bundle.crt", O_RDONLY) = 4 open("/etc/pki/tls/certs/ca-bundle.crt", O_RDONLY) = 4 Appending my CA's certificate to ca-bundle.crt works of course, but it's a much more fragile and less elegant solution. Does anyone know what the proper way is now? Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.00 0.04 0.11 From hbrown at divms.uiowa.edu Wed Feb 23 18:09:02 2011 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Wed, 23 Feb 2011 12:09:02 -0600 Subject: [rhelv6-list] Subversion client with https and internal CA on RHEL6 In-Reply-To: <20110223175910.053e059d@python3.es.aed.lan> References: <20110223175910.053e059d@python3.es.aed.lan> Message-ID: <4D654D3E.4060905@divms.uiowa.edu> On 02/23/2011 10:59 AM, Matthias Saou wrote: > Hi, > > I can't seem to figure out the proper clean way to have the svn CLI > client trust all https URLs using certificates signed by an internal > CA. With RHEL5, it was easy : > > $ strace svn co https://myserver/repo/ 2>&1 | grep pki > open("/etc/pki/tls/cert.pem", O_RDONLY) = 3 > open("/etc/pki/tls/cert.pem", O_RDONLY) = 3 > stat("/etc/pki/tls/certs/b903d65c.0", 0x7fff7f839980) = -1 ENOENT (No > such file or directory) > > The SSL library being used looked for a CA certificate named after a > hash specific to that certificate (which you got with openssl x509 > -hash -noout -in myca.crt). With RHEL6 this happens no more : > > $ strace svn co https://myserver/repo/ 2>&1 | grep pki > open("/etc/pki/tls/certs/ca-bundle.crt", O_RDONLY) = 4 > open("/etc/pki/tls/certs/ca-bundle.crt", O_RDONLY) = 4 > > Appending my CA's certificate to ca-bundle.crt works of course, but > it's a much more fragile and less elegant solution. Does anyone know > what the proper way is now? > > Matthias > In ~/.subversion/servers you can set ssl-authority-files to point at your CAs cert in pem format Hugh -- System Administrator University of Iowa DIVMS Support Group hbrown at divms.uiowa.edu Have a problem? Send mail to request at divms.uiowa.edu From gsgatlin at ncsu.edu Wed Feb 23 20:58:48 2011 From: gsgatlin at ncsu.edu (Gary Gatling) Date: Wed, 23 Feb 2011 15:58:48 -0500 (EST) Subject: [rhelv6-list] power saving (suspend) automatic setup Message-ID: Hello, I am trying to automate a RHEL 6 install that will produce a machine that suspends to RAM after 600 seconds. (10 minutes of inactivity) If I log into a freshly insatalled machine and run gnome-power-preferences, and then set "Put computer to sleep when inactive for :" 10 minutes and then log out, it seems the computer suspends to RAM as expected. (I know I've seen it do the right thing once at least) If I set these gconf strings, the computer does not suspend to RAM. maybe gnome-power-preferences is doing other things besides setting gconf strings? gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_ac 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_battery 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_ups 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_display_ac 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type bool --set /apps/gnome-power-manager/disks/spindown_enable_ac true Has anyone figured out a automated way to install a machine and have it sleep after a certian amount of time automatically? Wanting to do this to potentially save power costs in big computer labs. I am creating an rpm package to deliver the settings in a kickstart install. I thought about having a cron job just run: /usr/sbin/pm-suspend every 15 minutes if no one is logged in. But that won't work very well if someone is just logging in just as the machine runs the sleep cron. That might piss off some users and seem like a bug. Also, do I need to set these settings as the "gdm" user rather than root? I ask since gdm is running on X when no one is logged in. Thanks for any ideas anyone has? Gary Gatling | ITECS Systems From hbrown at divms.uiowa.edu Wed Feb 23 22:07:59 2011 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Wed, 23 Feb 2011 16:07:59 -0600 Subject: [rhelv6-list] Replacement for /etc/security/console.perms.d/50-default.perms? Message-ID: <4D65853F.6010406@divms.uiowa.edu> Under RHEL5, the 50-default.perms file set file permissions on a variety of things. I noticed its absence in the case of /dev/nvidia*. It used to be that whoever logged in at the console got 0600 and ownership. Now, they stay root:root and 0666. What replaced that functionality? I haven't found anything in udev. I'm assuming that any user who can log in can snoop my X session since rw access is available to everyone on /dev/nvidia* Hugh From gsgatlin at ncsu.edu Thu Feb 24 16:17:11 2011 From: gsgatlin at ncsu.edu (Gary Gatling) Date: Thu, 24 Feb 2011 11:17:11 -0500 (EST) Subject: [rhelv6-list] SOLVED power saving (suspend) automatic setup In-Reply-To: References: Message-ID: Hello, I kind of stumbled across the answer to my question yesterday by accident. I thought some other RHEL 6 users might find what I am trying to do interesting? (I'd like to do this on my home CentOS/fedora boxes at some point) It turns out when the machine first boots and no one has logged in yet "root" controls the power settings. But if root has never had an X session, the dotfiles will be missing. So I needed to run commands like: gconftool-2 --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_ac 600 as user root. But I didn't want to change the default settings for all users. If someone logs in and wants to change "their" settings they can in the preferences app but by default I want it not to hibernate or suspend. Only want to suspend when the machines are idle or unused. Here is what I am trying to do.. We have a system here called VCL. (http://vcl.ncsu.edu/) A student can go to a web page and request a machine so that they can ssh into it. Then they can run whatever programs they want back to their X display. The web page spits out an IP address for them to use. That way labs can be used even when they are physically closed by users at home. (The web app will pick the fastest systems in whatever labs) Gnome doesn't know about students sshing into a box. It will just suspend a box if someone is logged in through some other method than GNOME. :( So I run a shell script via a cron job every 15 minutes to see if someone is logged in. If they are, I set root's setting NOT to suspend. If no one is logged in I change root's settings TO suspend after 600 seconds of inactivity. This should keep it from suspending right on the hour mark mostly I hope in case I run wol scripts on an even hour via cron from a server. It seems to work as expected. One thing I discovered about suspend to RAM in Linux is that normally devices don't have permission to wake up a PC. (Like ethernet) So besides setting up stuff in the BIOS, you have to mess with the file /proc/acpi/wakeup, at least in RHEL 6. Probably older RHELs/fedoras also. /proc/acpi/wakeup contains a list of devices that could possibly wake up the pc. By default they are all disabled. I made a script to walk the list and change each / every entry to enabled. The script runs at each boot as a init script in runlevel 345. With these two things in place I can send "wol" packets to the PCs to wake them or a user can press the keyboard and wake the PC to log in. The mouse on my DELL doesn't seem to wake the PC for some reason. But on our KVM switch the mouse does wake the PC. I guess its some kind of power issue with usb probably. They can also press the power button. Eventually the script could be made fancier and do hibernate instead of suspend. Maybe if no one has logged in for like 3 days? If anyone else might be interested in the rpm I created to do this I could make it availible in a public repo. (Its very simple) I do need to tweak it some more before I can use it. The power saving improvements in RHEL 6 seem pretty cool to be able to do stuff like this. So thanks Red Hat. :) Cheers, Gary Gatling | ITECS Systems From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Feb 24 16:24:07 2011 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 24 Feb 2011 17:24:07 +0100 Subject: [rhelv6-list] Subversion client with https and internal CA on RHEL6 In-Reply-To: <4D654D3E.4060905@divms.uiowa.edu> References: <20110223175910.053e059d@python3.es.aed.lan> <4D654D3E.4060905@divms.uiowa.edu> Message-ID: <20110224172407.2938eff1@python3.es.aed.lan> Hugh Brown wrote : > In ~/.subversion/servers you can set ssl-authority-files to point at > your CAs cert in pem format I was kind of hoping to have my CA recognized on some system-wide scale, but that will have to do for now. Thanks for the pointer, it works fine :-) Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 14 (Laughlin) - Linux kernel 2.6.35.10-72.fc14.x86_64 Load : 0.00 0.02 0.05 From gsgatlin at eos.ncsu.edu Wed Feb 23 20:54:45 2011 From: gsgatlin at eos.ncsu.edu (Gary Gatling) Date: Wed, 23 Feb 2011 15:54:45 -0500 (EST) Subject: [rhelv6-list] power saving (suspend) automatic setup Message-ID: Hello, I am trying to automate a RHEL 6 install that will produce a machine that suspends to RAM after 600 seconds. (10 minutes of inactivity) If I log into a freshly insatalled machine and run gnome-power-preferences, and then set "Put computer to sleep when inactive for :" 10 minutes and then log out, it seems the computer suspends to RAM as expected. (I know I've seen it do the right thing once at least) If I set these gconf strings, the computer does not suspend to RAM. maybe gnome-power-preferences is doing other things besides setting gconf strings? gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_ac 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_battery 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_computer_ups 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type integer --set /apps/gnome-power-manager/timeout/sleep_display_ac 600 gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type bool --set /apps/gnome-power-manager/disks/spindown_enable_ac true Has anyone figured out a automated way to install a machine and have it sleep after a certian amount of time automatically? Wanting to do this to potentially save power costs in big computer labs. I am creating an rpm package to deliver the settings in a kickstart install. I thought about having a cron job just run: /usr/sbin/pm-suspend every 15 minutes if no one is logged in. But that won't work very well if someone is just logging in just as the machine runs the sleep cron. That might piss off some users and seem like a bug. Also, do I need to set these settings as the "gdm" user rather than root? I ask since gdm is running on X when no one is logged in. Thanks for any ideas anyone has? Gary Gatling | ITECS Systems