From lowen at pari.edu Thu Aug 11 15:18:26 2011 From: lowen at pari.edu (Lamar Owen) Date: Thu, 11 Aug 2011 11:18:26 -0400 Subject: [rhelv6-list] KDE greater than 4.3.4 Message-ID: <201108111118.26720.lowen@pari.edu> Ok, does anyone know of a repository or rebuild effort that has KDE packages for something more modern than 4.3.4? I've looked at kde-redhat, but the packages there are pretty old for EL, and none that I could see on apt.kde-redhat.org for EL6. I really prefer KDE to GNOME, but going from my Fedora 14 desktop to an RHEL 6.1 desktop with KDE is somewhat painful (my RHEL 6.1 box, while more of a server, is also running desktop stuff). (And I'm well aware that using such would be completely unsupported.....) And while KDE 4.3.4 was a good, stable, solid release, I have grown rather accustomed to 4.6.x on F14.... From robinprice at gmail.com Thu Aug 11 19:00:18 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Thu, 11 Aug 2011 15:00:18 -0400 Subject: [rhelv6-list] KDE greater than 4.3.4 In-Reply-To: <201108111118.26720.lowen@pari.edu> References: <201108111118.26720.lowen@pari.edu> Message-ID: Lamar, RHEL6 is still in it's early stages. I have seen other packages rebase to upstream version (firefox) before in a RHEL release. While I don't have the answer as to there being a repo to house the updated packages for RHEL6, it wouldn't heard to open a BZ asking to have it rebased to the latest. Any justifications to the rebase would greatly help with the engineering decision on this as well. If the justification is for an over all better desktop experience or functionality that is missing in the KDE RHEL community, it would help in the bugzilla. I would also like to invite you the RHEL community at: https://access.redhat.com/groups/red-hat-enterprise-linux It would be great to take this discussion there also and get other peoples opinions who are active in our Customer Support Portal and may not be on this list. There are a lot of Red Hat employee and engineers checking their daily. And if it makes you feel any better, I use RHEL6.1 on my laptop has a workstation but it's registered as a RHEL6 server ;) Works great! And I love the version of GNOME that is shipped with it. Hope this helps! ~rp On Thu, Aug 11, 2011 at 11:18 AM, Lamar Owen wrote: > Ok, does anyone know of a repository or rebuild effort that has KDE packages for something more modern than 4.3.4? ?I've looked at kde-redhat, but the packages there are pretty old for EL, and none that I could see on apt.kde-redhat.org for EL6. > > I really prefer KDE to GNOME, but going from my Fedora 14 desktop to an RHEL 6.1 desktop with KDE is somewhat painful (my RHEL 6.1 box, while more of a server, is also running desktop stuff). ?(And I'm well aware that using such would be completely unsupported.....) > > And while KDE 4.3.4 was a good, stable, solid release, I have grown rather accustomed to 4.6.x on F14.... > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From lowen at pari.edu Thu Aug 11 20:15:01 2011 From: lowen at pari.edu (Lamar Owen) Date: Thu, 11 Aug 2011 16:15:01 -0400 Subject: [rhelv6-list] KDE greater than 4.3.4 In-Reply-To: References: <201108111118.26720.lowen@pari.edu> Message-ID: <201108111615.01699.lowen@pari.edu> On Thursday, August 11, 2011 03:00:18 PM robinprice at gmail.com wrote: > RHEL6 is still in it's early stages. I have seen other packages > rebase to upstream version (firefox) before in a RHEL release. While > I don't have the answer as to there being a repo to house the updated > packages for RHEL6, it wouldn't heard to open a BZ asking to have it > rebased to the latest. Thanks for the reply, and for the online community discussion link. I've been a KDE user since Mandrake 5.3/RHL 6.0 days, and have a huge amount of e-mail and other information in Kontact that would be rather difficult to migrate to something else (~10GB or so in a few hundred folders). Never really have been a fan of GNOME, even when testing. Seemed too oversimplified for my tastes. And evolution is just as limiting as Outlook for the way I use e-mail, and for the productivity I currently have in kmail and the other kontact applications. But I understand that I'm not the typical user. I'll work up a BZ, and probably file it against kdebase, tomorrow or early next week. Too tired right now. From bda20 at cam.ac.uk Thu Aug 11 21:40:00 2011 From: bda20 at cam.ac.uk (Ben) Date: Thu, 11 Aug 2011 22:40:00 +0100 (BST) Subject: [rhelv6-list] KDE greater than 4.3.4 In-Reply-To: <201108111615.01699.lowen@pari.edu> References: <201108111118.26720.lowen@pari.edu> <201108111615.01699.lowen@pari.edu> Message-ID: On Thu, 11 Aug 2011, Lamar Owen wrote: > I've been a KDE user since Mandrake 5.3/RHL 6.0 days, and have a huge > amount of e-mail and other information in Kontact that would be rather > difficult to migrate to something else (~10GB or so in a few hundred > folders). > > Never really have been a fan of GNOME, even when testing. Seemed too > oversimplified for my tastes. And evolution is just as limiting as > Outlook for the way I use e-mail, and for the productivity I currently > have in kmail and the other kontact applications. But I understand that > I'm not the typical user. > > I'll work up a BZ, and probably file it against kdebase, tomorrow or early > next week. Too tired right now. I believe I filed a BZ about getting a more recent version of KDE into RHEL6 back when it was in beta. Let me look... Ah yes, here it is: https://bugzilla.redhat.com/show_bug.cgi?id=609662 CLOSED, tagged as WONTFIX. It _might_ make it into RHEL7. Or perhaps by then it'll be KDE5.x they go with (-: Sorry, but unless something's changed I'd say your chances are somewhere between slim and none, and I think it missed dinner. You're not the only disappointed person. Ben -- Unix Support, MISD, University of Cambridge, England Plugger of wire, typer of keyboard, imparter of Clue Life Is Short. It's All Good. From vincent at cojot.name Fri Aug 12 08:54:03 2011 From: vincent at cojot.name (vincent at cojot.name) Date: Fri, 12 Aug 2011 10:54:03 +0200 (CEST) Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: <201108111118.26720.lowen@pari.edu> References: <201108111118.26720.lowen@pari.edu> Message-ID: Hi everyone, Speaking of desktop environments, what's the proper, clean, recommended way of adding a 3rd party desktop environment to RHEL6 without hacking into all of the /etc/gdm stuff? I cannot seem to find it in the documentation anywhere (gdm 2.30.x) and other people with similar needs seem to be hacking away under /etc/gdm. Under RHEL4/5, it was just a matter of adding stuff under /etc/X11/* as in: /etc/X11/dm/Sessions/OpenWindows.desktop /etc/X11/gdm/Sessions/OpenWin Any ideas/hints? Thanks in advance, Best regards, Vincent From philipdurbin at gmail.com Fri Aug 12 13:47:12 2011 From: philipdurbin at gmail.com (Philip Durbin) Date: Fri, 12 Aug 2011 09:47:12 -0400 Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: References: <201108111118.26720.lowen@pari.edu> Message-ID: <4E452EE0.3020908@gmail.com> On 08/12/2011 04:54 AM, vincent at cojot.name wrote: > Speaking of desktop environments, what's the proper, clean, recommended > way of adding a 3rd party desktop environment to RHEL6 without hacking > into all of the /etc/gdm stuff? I cannot seem to find it in the > documentation anywhere (gdm 2.30.x) and other people with similar needs > seem to be hacking away under /etc/gdm. > > Under RHEL4/5, it was just a matter of adding stuff under /etc/X11/* as in: > /etc/X11/dm/Sessions/OpenWindows.desktop > /etc/X11/gdm/Sessions/OpenWin For example: /usr/share/applications/gnome-terminal.desktop Phil From vincent at cojot.name Fri Aug 12 14:44:16 2011 From: vincent at cojot.name (vincent at cojot.name) Date: Fri, 12 Aug 2011 16:44:16 +0200 (CEST) Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: <4E452EE0.3020908@gmail.com> References: <201108111118.26720.lowen@pari.edu> <4E452EE0.3020908@gmail.com> Message-ID: On Fri, 12 Aug 2011, Philip Durbin wrote: >> Under RHEL4/5, it was just a matter of adding stuff under /etc/X11/* as in: >> /etc/X11/dm/Sessions/OpenWindows.desktop >> /etc/X11/gdm/Sessions/OpenWin > > For example: /usr/share/applications/gnome-terminal.desktop > > Phil > Hi Phil, Unless I am mistaken, stuff under /usr/share/applications is for gnome-compliant applications, not for a complete session chosen from GDM to be neither Gnome or KDE (which is what I'd like to do). I'll continue my readings.. Cheers, Vincent From hbrown at divms.uiowa.edu Fri Aug 12 15:13:14 2011 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Fri, 12 Aug 2011 10:13:14 -0500 Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: References: <201108111118.26720.lowen@pari.edu> <4E452EE0.3020908@gmail.com> Message-ID: <4E45430A.5080000@divms.uiowa.edu> On 08/12/2011 09:44 AM, vincent at cojot.name wrote: > On Fri, 12 Aug 2011, Philip Durbin wrote: > >>> Under RHEL4/5, it was just a matter of adding stuff under /etc/X11/* >>> as in: >>> /etc/X11/dm/Sessions/OpenWindows.desktop >>> /etc/X11/gdm/Sessions/OpenWin >> >> For example: /usr/share/applications/gnome-terminal.desktop >> >> Phil >> > > Hi Phil, > > Unless I am mistaken, stuff under /usr/share/applications is for > gnome-compliant applications, not for a complete session chosen from GDM > to be neither Gnome or KDE (which is what I'd like to do). > I'll continue my readings.. > > Cheers, > > Vincent On Fedora 14/15, it looks like you add a .desktop file to /usr/share/xsessions/ . I don't have a RH6 box handy that has any desktop environments on it, but that may help as a next place to start looking/researching. For xfce on a Fedora 15 box, a brief version of the xfce.desktop file is: [Desktop Entry] Version=1.0 Name=Xfce Session Comment=Use this session to run Xfce as your desktop environment Exec=startxfce4 Icon= Type=Application Hugh From vincent at cojot.name Fri Aug 12 15:37:15 2011 From: vincent at cojot.name (vincent at cojot.name) Date: Fri, 12 Aug 2011 17:37:15 +0200 (CEST) Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: <4E45430A.5080000@divms.uiowa.edu> References: <201108111118.26720.lowen@pari.edu> <4E452EE0.3020908@gmail.com> <4E45430A.5080000@divms.uiowa.edu> Message-ID: Hi Hugh, Well then I wonder if I should file a bug against that GDM since I already had something similar on my test RHEL6.1 system: [root at rh6x64 xsessions]# pwd /usr/share/xsessions [root at rh6x64 xsessions]# cat OpenWindows.desktop [Desktop Entry] Encoding=UTF-8 Type=XSession Exec=/usr/openwin/bin/openwin TryExec=/usr/openwin/bin/openwin Name=OpenWindows Comment=The OpenWindows Desktop But anyway I didn't see any place in the GDM greeter where to pick a different session that the default.. Vincent On Fri, 12 Aug 2011, Hugh Brown wrote: > On Fedora 14/15, it looks like you add a .desktop file to > /usr/share/xsessions/ . > > I don't have a RH6 box handy that has any desktop environments on it, but > that may help as a next place to start looking/researching. > > For xfce on a Fedora 15 box, a brief version of the xfce.desktop file is: > > > [Desktop Entry] > Version=1.0 > Name=Xfce Session > Comment=Use this session to run Xfce as your desktop environment > Exec=startxfce4 > Icon= > Type=Application > > > Hugh From hbrown at divms.uiowa.edu Fri Aug 12 16:01:43 2011 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Fri, 12 Aug 2011 11:01:43 -0500 Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: References: <201108111118.26720.lowen@pari.edu> <4E452EE0.3020908@gmail.com> <4E45430A.5080000@divms.uiowa.edu> Message-ID: <4E454E67.1060408@divms.uiowa.edu> On 08/12/2011 10:37 AM, vincent at cojot.name wrote: > > Hi Hugh, > > Well then I wonder if I should file a bug against that GDM since I > already had something similar on my test RHEL6.1 system: > > [root at rh6x64 xsessions]# pwd > /usr/share/xsessions > > [root at rh6x64 xsessions]# cat OpenWindows.desktop > [Desktop Entry] > Encoding=UTF-8 > Type=XSession > Exec=/usr/openwin/bin/openwin > TryExec=/usr/openwin/bin/openwin > Name=OpenWindows > Comment=The OpenWindows Desktop > > But anyway I didn't see any place in the GDM greeter where to pick a > different session that the default.. > > Vincent Does the gdm user have read access to the file? It's also possible that update-desktop-database may have to be run (though I don't know if it touches the desktop files in that directory). Hugh From hbrown at divms.uiowa.edu Fri Aug 12 16:04:36 2011 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Fri, 12 Aug 2011 11:04:36 -0500 Subject: [rhelv6-list] Proper way to add to RHEL6's GDM? In-Reply-To: References: <201108111118.26720.lowen@pari.edu> <4E452EE0.3020908@gmail.com> <4E45430A.5080000@divms.uiowa.edu> Message-ID: <4E454F14.4060006@divms.uiowa.edu> On 08/12/2011 10:37 AM, vincent at cojot.name wrote: > > Hi Hugh, > > Well then I wonder if I should file a bug against that GDM since I > already had something similar on my test RHEL6.1 system: > > [root at rh6x64 xsessions]# pwd > /usr/share/xsessions > > [root at rh6x64 xsessions]# cat OpenWindows.desktop > [Desktop Entry] > Encoding=UTF-8 > Type=XSession > Exec=/usr/openwin/bin/openwin > TryExec=/usr/openwin/bin/openwin > Name=OpenWindows > Comment=The OpenWindows Desktop > > But anyway I didn't see any place in the GDM greeter where to pick a > different session that the default.. > > Vincent I also notice that some of the desktop files have Type=Application and some have Type=XSession. I'm not sure what the implications are of that difference either. Hugh From evilensky at gmail.com Tue Aug 16 21:06:50 2011 From: evilensky at gmail.com (Eugene Vilensky) Date: Tue, 16 Aug 2011 16:06:50 -0500 Subject: [rhelv6-list] Host hung, hung_task_timeout_secs mentioned In-Reply-To: <1308153211.9779.16.camel@iso-8590-lx> References: <4DF8CC0D.9090307@cisco.com> <1308153211.9779.16.camel@iso-8590-lx> Message-ID: On Wed, Jun 15, 2011 at 10:53 AM, Tom Sightler wrote: > On Wed, 2011-06-15 at 11:13 -0400, Brian Long wrote: >> I ran into a server hang last night running 2.6.32-131.2.1.el6.x86_64. >> I just installed the latest updates (RHEL 6.1) yesterday morning and I >> experienced the hang during my Amanda backups. ?I found a RHEL 5 bug >> which mention similar problems but no fix: >> https://bugzilla.redhat.com/show_bug.cgi?id=605444 >> >> I had Opsware monitoring the host and it went offline completely for >> about 1 minute. ?Has anyone else experienced this? ?I'm running a LSI >> 8708EM2 RAID controller with battery-backed cache. >> > > We battled a very similar issue on one of our older systems that was > recently upgraded. ?Specifically the system was a Dell 2950 that servers > as a central backup server. ?This server runs NFS and Samba to take > Oracle RMAN backups, runs BackupPC to backup a number of Linux systems, > and is a backup target for our VMware backup solution. > > During the heavy I/O (somewhat common for a backup server) we would get > messages similar to what you're seeing. ?Interestingly, we have another > older server that performs virtually the same function but so far it > hasn't experienced this issue. We ran into this on a RHEL 5.7 server, using ext4 on a LVM-based volume backed by an Equallogic iscsi device. These threads were hung: dmesg | grep INFO: INFO: task pdflush:30278 blocked for more than 120 seconds. INFO: task jbd2/dm-10-8:2788 blocked for more than 120 seconds. INFO: task httpd:4058 blocked for more than 120 seconds. INFO: task httpd:4278 blocked for more than 120 seconds. INFO: task httpd:12351 blocked for more than 120 seconds. INFO: task sftp-server:19531 blocked for more than 120 seconds. INFO: task sftp-server:19561 blocked for more than 120 seconds. INFO: task sftp-server:19562 blocked for more than 120 seconds. INFO: task sftp-server:19613 blocked for more than 120 seconds. INFO: task sftp-server:19637 blocked for more than 120 seconds. Not sure if it's IO load related, but ext4 sure throws up some warnings: EXT4-fs (dm-10): delayed block allocation failed for inode 15122461 at logical offset 0 with max blocks 1 with error -122 This should not happen!! Data will be lost [etc etc] From marco.shaw at gmail.com Tue Aug 16 23:40:47 2011 From: marco.shaw at gmail.com (Marco Shaw) Date: Tue, 16 Aug 2011 20:40:47 -0300 Subject: [rhelv6-list] /etc/sysconfig/network Message-ID: RE: /etc/sysconfig/network Did a quick search and came up empty... Is this file still used in RHEL6? From what I can remember, it was included in RHEL5, but must be manually created on RHEL6? Marco From colin.coe at gmail.com Wed Aug 17 01:10:49 2011 From: colin.coe at gmail.com (Colin Coe) Date: Wed, 17 Aug 2011 09:10:49 +0800 Subject: [rhelv6-list] /etc/sysconfig/network In-Reply-To: References: Message-ID: I spun up a new RHEL6 box about 5 minutes ago and it had this file although only 3 lines including the GATEWAY line which I commented out as it was already in the ifcfg-eth0 file. NETWORKING=yes HOSTNAME=blah #GATEWAY=x.x.x.x CC On Wed, Aug 17, 2011 at 7:40 AM, Marco Shaw wrote: > RE: /etc/sysconfig/network > > Did a quick search and came up empty... ?Is this file still used in > RHEL6? ?From what I can remember, it was included in RHEL5, but must > be manually created on RHEL6? > > Marco > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > -- RHCE#805007969328369 From digital at pghfilmmakers.org Wed Aug 17 01:37:43 2011 From: digital at pghfilmmakers.org (Pittsburgh Filmmakers Institute) Date: Tue, 16 Aug 2011 21:37:43 -0400 Subject: [rhelv6-list] /etc/sysconfig/network In-Reply-To: References: Message-ID: <201108162137.43130.digital@pghfilmmakers.org> Yeah you still need it. For example: [wheelishuser at rhel6]$ cat /etc/sysconfig/network-scripts/ifcfg-eth0 DEVICE=eth0 NM_CONTROLLED=yes ONBOOT=yes TYPE=Ethernet BOOTPROTO=none IPADDR=69.156.21.46 PREFIX=24 GATEWAY=69.156.21.1 NETMASK=255.255.255.0 DNS1=192.168.0.3 DNS2=208.67.222.222 PEERDNS=no DOMAIN=pghfilmmakers.local SEARCH=pghfilmmakers.local DEFROUTE=yes IPV4_FAILURE_FATAL=yes IPV6INIT=no NAME="System eth0" HWADDR=1d:de:1c:0a:ef:de USERCTL=no From john.haxby at gmail.com Wed Aug 17 08:19:04 2011 From: john.haxby at gmail.com (John Haxby) Date: Wed, 17 Aug 2011 09:19:04 +0100 Subject: [rhelv6-list] /etc/sysconfig/network In-Reply-To: References: Message-ID: On 17 August 2011 00:40, Marco Shaw wrote: > > Did a quick search and came up empty... Is this file still used in > RHEL6? From what I can remember, it was included in RHEL5, but must > be manually created on RHEL6? > > I suspect, but haven't checked, that it's not created if you're just using dhcp and NetworkManager. jch -------------- next part -------------- An HTML attachment was scrubbed... URL: From bluetux at gmail.com Fri Aug 19 02:30:55 2011 From: bluetux at gmail.com (Kitae Kim) Date: Fri, 19 Aug 2011 11:30:55 +0900 Subject: [rhelv6-list] Question about fence device network? Message-ID: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> Hello, I try to build Redhat H/A system, following Cluster Administration guide. My Question is about fencing network. (ex, HP iLo) Some Engineer said fencing network, should use gigabit switch. I think more effective system fencing network should separated from service netwrok. Fencing network is very import for stable service and control H/A system, but I think 100Mb/s is enough speed for fencing network, isn't it? or should I use 1 Gigabit network? Thanks ------------------ Creative & Positive Mind From greg at nytefyre.net Fri Aug 19 02:40:56 2011 From: greg at nytefyre.net (Greg Swift) Date: Thu, 18 Aug 2011 21:40:56 -0500 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> Message-ID: the throughput of that network shouldn't matter, ecspecially if its a dedicated network. On 2011-08-18, Kitae Kim wrote: > Hello, > > I try to build Redhat H/A system, following Cluster Administration guide. > > My Question is about fencing network. (ex, HP iLo) > > Some Engineer said fencing network, should use gigabit switch. > I think more effective system fencing network should separated from service > netwrok. > > Fencing network is very import for stable service and control H/A system, > but > I think 100Mb/s is enough speed for fencing network, isn't it? > or should I use 1 Gigabit network? > > Thanks > > ------------------ > Creative & Positive Mind > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From linux at alteeve.com Fri Aug 19 03:18:48 2011 From: linux at alteeve.com (Digimer) Date: Thu, 18 Aug 2011 23:18:48 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> Message-ID: <4E4DD618.3000906@alteeve.com> On 08/18/2011 10:30 PM, Kitae Kim wrote: > Hello, > > I try to build Redhat H/A system, following Cluster Administration guide. > > My Question is about fencing network. (ex, HP iLo) > > Some Engineer said fencing network, should use gigabit switch. > I think more effective system fencing network should separated from service netwrok. > > Fencing network is very import for stable service and control H/A system, but > I think 100Mb/s is enough speed for fencing network, isn't it? > or should I use 1 Gigabit network? > > Thanks The speed of the network doesn't really matter. It's useful to have the fencing on a *separate* network, as it would allow for the fencing of a node (and thus allow recovery to begin) when there is a network error that breaks cluster communications. This is mainly a 2-node issue though. -- Digimer E-Mail: digimer at alteeve.com Freenode handle: digimer Papers and Projects: http://alteeve.com Node Assassin: http://nodeassassin.org "At what point did we forget that the Space Shuttle was, essentially, a program that strapped human beings to an explosion and tried to stab through the sky with fire and math?" From solarflow99 at gmail.com Fri Aug 19 04:55:00 2011 From: solarflow99 at gmail.com (solarflow99) Date: Fri, 19 Aug 2011 00:55:00 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: <4E4DD618.3000906@alteeve.com> References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> Message-ID: i'm going through the guide too and I just wonder why you would even need fencing if you are not using something like GFS? Imagine a network with many app servers running on their own storage, they would probably just link to a database to read/write any data. -------------- next part -------------- An HTML attachment was scrubbed... URL: From linux at alteeve.com Fri Aug 19 06:12:52 2011 From: linux at alteeve.com (Digimer) Date: Fri, 19 Aug 2011 02:12:52 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> Message-ID: <4E4DFEE4.9060406@alteeve.com> On 08/19/2011 12:55 AM, solarflow99 wrote: > i'm going through the guide too and I just wonder why you would even > need fencing if you are not using something like GFS? Imagine a network > with many app servers running on their own storage, they would probably > just link to a database to read/write any data. The cluster communication and membership layer is separate from the resource management layer. The membership layer is concerned with having nodes in a known state. So, when a node "disappears" (network failure, hard failure, kernel panic, hung/blocked), the node falls into an unknown state. The cluster blocks at will remain blocked until the lost node is put into a known state. This is done using fencing. By allowing the cluster to reliably reach out and force-off a node, the node then is in a known state and recovery can begin. All of this happens well below the resource manager, and is not dependant on whether something like GFS2 is in use or not. -- Digimer E-Mail: digimer at alteeve.com Freenode handle: digimer Papers and Projects: http://alteeve.com Node Assassin: http://nodeassassin.org "At what point did we forget that the Space Shuttle was, essentially, a program that strapped human beings to an explosion and tried to stab through the sky with fire and math?" From RJM002 at SHSU.EDU Fri Aug 19 11:30:04 2011 From: RJM002 at SHSU.EDU (Marti, Robert) Date: Fri, 19 Aug 2011 06:30:04 -0500 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> Message-ID: <1B116D93-4A13-4803-A3CB-56DF476698BF@shsu.edu> The purpose of fencing is to ensure a cluster resource isn't ruined by a member in an unknown state. If you're not using a cluster resource (no shared storage, no failover applications) why are you running cluster software? On Aug 18, 2011, at 23:57, "solarflow99" wrote: > i'm going through the guide too and I just wonder why you would even need fencing if you are not using something like GFS? Imagine a network with many app servers running on their own storage, they would probably just link to a database to read/write any data. > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list From solarflow99 at gmail.com Fri Aug 19 21:31:34 2011 From: solarflow99 at gmail.com (solarflow99) Date: Fri, 19 Aug 2011 17:31:34 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: <1B116D93-4A13-4803-A3CB-56DF476698BF@shsu.edu> References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> <1B116D93-4A13-4803-A3CB-56DF476698BF@shsu.edu> Message-ID: I see, it just makes me wonder a few things. Can HA cluster load balance traffic on each node? I didn't see anything about that, most clusters would want to have other nodes busy serving requests. Or is it supposed to work like this: use LVS to load balance to the app servers, and use HA cluster for shared NFS storage? I can't see how something like a database can read and write like crazy like this without performance problems. Thanks, -------------- next part -------------- An HTML attachment was scrubbed... URL: From linux at alteeve.com Fri Aug 19 21:38:44 2011 From: linux at alteeve.com (Digimer) Date: Fri, 19 Aug 2011 17:38:44 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> <1B116D93-4A13-4803-A3CB-56DF476698BF@shsu.edu> Message-ID: <4E4ED7E4.60109@alteeve.com> On 08/19/2011 05:31 PM, solarflow99 wrote: > I see, it just makes me wonder a few things. Can HA cluster load > balance traffic on each node? I didn't see anything about that, most > clusters would want to have other nodes busy serving requests. Or is it > supposed to work like this: use LVS to load balance to the app servers, > and use HA cluster for shared NFS storage? I can't see how something > like a database can read and write like crazy like this without > performance problems. > > Thanks, "Cluster" simply means "two or more servers working together". Nothing more. Now that said, load-balancing is a very common form performance clustering, and they inherently offer a degree of high availability. That is am implementation choice though, above and beyond the cluster foundation itself. Consider; Another common cluster is one that hosts high-availability virtual servers. If you have only one VM, then the second node is sitting there doing nothing. This is still very much a valid cluster. Another example would be an HA SAN with a floating IP. The backup is just sitting there, effectively doing nothing, but again, this is a perfectly valid cluster. You need to make a mental separation between the cluster foundation and the cluster resources. They work intimately together, but they are fundamentally separate components. -- Digimer E-Mail: digimer at alteeve.com Freenode handle: digimer Papers and Projects: http://alteeve.com Node Assassin: http://nodeassassin.org "At what point did we forget that the Space Shuttle was, essentially, a program that strapped human beings to an explosion and tried to stab through the sky with fire and math?" From solarflow99 at gmail.com Mon Aug 22 07:23:33 2011 From: solarflow99 at gmail.com (solarflow99) Date: Mon, 22 Aug 2011 03:23:33 -0400 Subject: [rhelv6-list] Question about fence device network? In-Reply-To: <4E4ED7E4.60109@alteeve.com> References: <3F8A71D4-219D-497F-8D73-B31FA629F4D7@gmail.com> <4E4DD618.3000906@alteeve.com> <1B116D93-4A13-4803-A3CB-56DF476698BF@shsu.edu> <4E4ED7E4.60109@alteeve.com> Message-ID: I just wonder what most others prefer for database high availability? From this link: http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Load_Balancer_Administration/ch-clumanager-piranha-VSA.html#fig-lvs-rhclust-VSA It describes a third tier, but with ssd's replacing the need for a SAN and GFS, is there a better way to be doing this? -------------- next part -------------- An HTML attachment was scrubbed... URL: From colin.coe at gmail.com Fri Aug 26 05:44:36 2011 From: colin.coe at gmail.com (Colin Coe) Date: Fri, 26 Aug 2011 13:44:36 +0800 Subject: [rhelv6-list] iSCSI on HP Flex10 interconnects Message-ID: Hi all Is anyone out there successfully doing _hardware_ iSCSI on HP Blade services through a Flex10 interconnect using RHEL6? I'm breaking my brain trying to make this work. Thanks CC PS: Yep, I know I've given no detail. I'll expand on this if I get any responses. -- RHCE#805007969328369 From domain-nospam at abdussamad.com Mon Aug 29 12:33:31 2011 From: domain-nospam at abdussamad.com (Abdussamad) Date: Mon, 29 Aug 2011 12:33:31 +0000 (UTC) Subject: [rhelv6-list] Kernel memory leak? References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: Anyone fix this? I have the same problem on my Centos 6 server: Active / Total Caches (% used) : 94 / 143 (65.7%) Active / Total Size (% used) : 280569.70K / 290010.73K (96.7%) Minimum / Average / Maximum Object : 0.02K / 0.22K / 128.00K OBJS ACTIVE USE OBJ SIZE SLABS OBJ/SLAB CACHE SIZE NAME 809964 809964 100% 0.21K 44998 18 179992K dentry_cache 288200 252787 87% 0.09K 7205 40 28820K buffer_head 80205 79947 99% 0.74K 16041 5 64164K ext3_inode_cache 28364 28286 99% 0.52K 4052 7 16208K radix_tree_node 19747 5667 28% 0.30K 1519 13 6076K ip_conntrack 15753 7966 50% 0.06K 267 59 1068K size-64 9394 8421 89% 0.17K 427 22 1708K vm_area_struct 6660 6053 90% 0.12K 222 30 888K size-128 5588 5565 99% 0.09K 127 44 508K sysfs_dir_cache 3045 2925 96% 0.25K 203 15 812K size-256 2576 2305 89% 0.03K 23 112 92K size-32 1440 809 56% 0.02K 10 144 40K anon_vma 1425 648 45% 0.25K 95 15 380K filp 1416 1303 92% 0.06K 24 59 96K Acpi-Operand 1204 1084 90% 0.55K 172 7 688K inode_cache 1104 862 78% 0.08K 23 48 92K selinux_inode_security 638 630 98% 2.00K 319 2 1276K size-2048 560 539 96% 0.03K 5 112 20K Acpi-Namespace 525 277 52% 0.25K 35 15 140K skbuff_head_cache 516 490 94% 0.58K 86 6 344K proc_inode_cache 496 468 94% 0.50K 62 8 248K size-512 400 380 95% 1.00K 100 4 400K size-1024 354 142 40% 0.06K 6 59 24K pid 354 139 39% 0.06K 6 59 24K delayacct_cache 354 33 9% 0.06K 6 59 24K blkdev_ioc 325 306 94% 0.75K 65 5 260K shmem_inode_cache 300 264 88% 0.12K 10 30 40K bio Write failed: Broken pipe From dsavage at peaknet.net Mon Aug 29 13:29:02 2011 From: dsavage at peaknet.net (Robert G. (Doc) Savage) Date: Mon, 29 Aug 2011 08:29:02 -0500 Subject: [rhelv6-list] GNOME Terminal dimensions Message-ID: <1314624542.3391.6.camel@lion.protogeek.org> In RHEL5, GNOME Terminal displayed a small X-Y pop-up while it was being resized. It's not there in RHEL6. Is this a configurable feature? If so, where? --Robert G. (Doc) Savage Fairview Heights, IL -------------- next part -------------- An HTML attachment was scrubbed... URL: From robinprice at gmail.com Mon Aug 29 21:53:09 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Mon, 29 Aug 2011 17:53:09 -0400 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <20110121164942.GG15266@hiwaay.net> <20110122040654.GA5036@hiwaay.net> Message-ID: I'll check around, but I haven't heard anything about RHEL6 and a memory leak issue. Then again, I am using RHEL6.1 , but looking around BZ and kbase, I can't find anything. It doesn't sound more like a [1] "Linux ate my RAM" issue more than anything. The following kbase is a nice overview to give you a better idea. https://access.redhat.com/kb/docs/DOC-52712 ~rp [1] http://www.linuxatemyram.com/ On Mon, Aug 29, 2011 at 8:33 AM, Abdussamad wrote: > Anyone fix this? I have the same problem on my Centos 6 server: > > > > Active / Total Caches (% used) ? ? : 94 / 143 (65.7%) > ?Active / Total Size (% used) ? ? ? : 280569.70K / 290010.73K (96.7%) > ?Minimum / Average / Maximum Object : 0.02K / 0.22K / 128.00K > > ?OBJS ACTIVE ?USE OBJ SIZE ?SLABS OBJ/SLAB CACHE SIZE NAME > 809964 809964 100% ? ?0.21K ?44998 ? ? ? 18 ? ?179992K dentry_cache > 288200 252787 ?87% ? ?0.09K ? 7205 ? ? ? 40 ? ? 28820K buffer_head > ?80205 ?79947 ?99% ? ?0.74K ?16041 ? ? ? ?5 ? ? 64164K ext3_inode_cache > ?28364 28286 ?99% ? ?0.52K ? 4052 ? ? ? ?7 ? ? 16208K radix_tree_node > ?19747 ? 5667 ?28% ? ?0.30K ? 1519 ? ? ? 13 ? ? ?6076K ip_conntrack > ?15753 7966 50% ? ?0.06K ? ?267 ? ? ? 59 ? ? ?1068K size-64 > ?9394 8421 89% ? ?0.17K ? ?427 ? ? ? 22 ? ? ?1708K vm_area_struct > ?6660 ? 6053 ?90% ? ?0.12K ? ?222 ? ? ? 30 ? ? ? 888K size-128 > ?5588 ? 5565 ?99% ? ?0.09K ? ?127 ? ? ? 44 ? ? ? 508K sysfs_dir_cache > ?3045 2925 96% ? ?0.25K ? ?203 ? ? ? 15 ? ? ? 812K size-256 > ?2576 ? 2305 ?89% ? ?0.03K ? ? 23 ? ? ?112 ? ? ? ?92K size-32 > ?1440 ? ?809 ?56% ? ?0.02K ? ? 10 ? ? ?144 ? ? ? ?40K anon_vma > ?1425 ? ?648 ?45% ? ?0.25K ? ? 95 ? ? ? 15 ? ? ? 380K filp > ?1416 ? 1303 ?92% ? ?0.06K ? ? 24 ? ? ? 59 ? ? ? ?96K Acpi-Operand > ?1204 ? 1084 ?90% ? ?0.55K ? ?172 ? ? ? ?7 ? ? ? 688K inode_cache > ?1104 ? ?862 ?78% ? ?0.08K ? ? 23 ? ? ? 48 ? ? ? ?92K selinux_inode_security > ? 638 ? ?630 ?98% ? ?2.00K ? ?319 ? ? ? ?2 ? ? ?1276K size-2048 > ? 560 ? ?539 ?96% ? ?0.03K ? ? ?5 ? ? ?112 ? ? ? ?20K Acpi-Namespace > ? 525 ? ?277 ?52% ? ?0.25K ? ? 35 ? ? ? 15 ? ? ? 140K skbuff_head_cache > ? 516 ? ?490 ?94% ? ?0.58K ? ? 86 ? ? ? ?6 ? ? ? 344K proc_inode_cache > ? 496 ? ?468 ?94% ? ?0.50K ? ? 62 ? ? ? ?8 ? ? ? 248K size-512 > ? 400 ? ?380 ?95% ? ?1.00K ? ?100 ? ? ? ?4 ? ? ? 400K size-1024 > ? 354 ? ?142 ?40% ? ?0.06K ? ? ?6 ? ? ? 59 ? ? ? ?24K pid > ? 354 ? ?139 ?39% ? ?0.06K ? ? ?6 ? ? ? 59 ? ? ? ?24K delayacct_cache > ? 354 ? ? 33 ? 9% ? ?0.06K ? ? ?6 ? ? ? 59 ? ? ? ?24K blkdev_ioc > ? 325 ? ?306 ?94% ? ?0.75K ? ? 65 ? ? ? ?5 ? ? ? 260K shmem_inode_cache > ? 300 ? ?264 ?88% ? ?0.12K ? ? 10 ? ? ? 30 ? ? ? ?40K bio > Write failed: Broken pipe > > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From robinprice at gmail.com Mon Aug 29 22:15:14 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Mon, 29 Aug 2011 18:15:14 -0400 Subject: [rhelv6-list] GNOME Terminal dimensions In-Reply-To: <1314624542.3391.6.camel@lion.protogeek.org> References: <1314624542.3391.6.camel@lion.protogeek.org> Message-ID: Robert, This shows for me. Are you using desktop effects by any chance? If you are, then I think it will not display, but you can try setting /apps/metacity/general/reduced_resources to true using the gconf-editor to get the old school grid and X-Y popup. ~rp On Mon, Aug 29, 2011 at 9:29 AM, Robert G. (Doc) Savage wrote: > In RHEL5, GNOME Terminal displayed a small X-Y pop-up while it was being > resized. It's not there in RHEL6. Is this a configurable feature? If so, > where? > > --Robert G. (Doc) Savage > ? Fairview Heights, IL > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > > From abdussamad at abdussamad.com Mon Aug 29 22:18:57 2011 From: abdussamad at abdussamad.com (Abdussamad Abdurrazzaq) Date: Tue, 30 Aug 2011 03:18:57 +0500 Subject: [rhelv6-list] Kernel memory leak? Message-ID: <4E5C1051.7070207@abdussamad.com> Hello Ok please ignore my previous email (if you've seen it). It's quite confused because I posted using gmane.org. I know about how Linux reports memory usage. My problem is very much real. Memory usage keeps increasing because of a memory leak in the kernel dentry cache. This is the same problem as outlined by others here: https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html So I was wondering whether this problem was fixed? I am using centos 6 with the following kernel: Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun 27 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux At one point dentry was using 3GB plus on my 8GB system! I am currently using a cron job to clear the cache every so often: sync && echo 2 >/proc/sys/vm/drop_caches The above works but I am looking for a more permanent solution. To that end I tried increasing: echo 10000> /proc/sys/vm/vfs_cache_pressure And in /etc/sysctl.conf But to no effect. So any idea how to fix this? Regards, Abdussamad From robinprice at gmail.com Mon Aug 29 23:39:27 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Mon, 29 Aug 2011 19:39:27 -0400 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5C1051.7070207@abdussamad.com> References: <4E5C1051.7070207@abdussamad.com> Message-ID: On Mon, Aug 29, 2011 at 6:18 PM, Abdussamad Abdurrazzaq wrote: > Hello > > Ok please ignore my previous email (if you've seen it). It's quite confused > because I posted using gmane.org. > > I know about how Linux reports memory usage. My problem is very much real. > Memory usage keeps increasing because of a memory leak in the kernel dentry > cache. This is the same problem as outlined by others here: > > https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html > > So I was wondering whether this problem was fixed? I am using centos 6 with > the following kernel: > > Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun 27 > 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux > > At one point dentry was using 3GB plus on ?my 8GB system! > > I am currently using a cron job to clear the cache every so often: > > sync && echo 2 >/proc/sys/vm/drop_caches > > The above works but I am looking for a more permanent solution. To that end > I tried increasing: > > echo 10000> /proc/sys/vm/vfs_cache_pressure > > And in /etc/sysctl.conf But to no effect. > > So any idea how to fix this? > > Regards, > Abdussamad > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > There are some things you can try to do. You can collect a vmcore from a time period during which the system has exhausted nearly all of it's memory due to this leak. Then you could try to analyze kmem to indicate if there is a problem with the kernel. You could even go as simple as just looking at top, or the contents of /proc//status periodically for the set of apps you suspect. If you do have a single app that is leaking memory, you should be able to record and graph a consistent increasing trend in the amount of memory the faulty app is leaking. That would at least give you a starting point for where to use an app like valgrind. Also, if you don't know which proc is at fault, I'd start with /etc/crontab: */5 * * * * root ps axo comm,vsize,rss | tail -n +2 >> /tmp/rawdata This would collect /proc/pid/statm data for a while. This should work on an selinux-enforcing machine based on the output of: # sesearch -As crond_t | grep tmp Then use awk to find the low- and high-water marks for each comm. You could get fancy and add timestamps to the records; maybe track current, low-water, and high-water marks, then gnuplot with error bars. As far as your drop_caches work around, know that drop_caches may cause performance degrade because some cached data are flushed and system have to load them from disk if they are needed again. Use the "ps aux" above results in the cron job and locate which program have a growing RSS. And sysstat (/var/log/sa/sar*) may provide some historical memory information that you may be interested in. Hope this gives you somewhere to look. I will follow-up on the thread you mentioned. ~rp From abdussamad at abdussamad.com Mon Aug 29 23:52:04 2011 From: abdussamad at abdussamad.com (Abdussamad Abdurrazzaq) Date: Tue, 30 Aug 2011 04:52:04 +0500 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <4E5C1051.7070207@abdussamad.com> Message-ID: <4E5C2624.8070903@abdussamad.com> On 08/30/2011 04:39 AM, robinprice at gmail.com wrote: > On Mon, Aug 29, 2011 at 6:18 PM, Abdussamad Abdurrazzaq > wrote: >> Hello >> >> Ok please ignore my previous email (if you've seen it). It's quite confused >> because I posted using gmane.org. >> >> I know about how Linux reports memory usage. My problem is very much real. >> Memory usage keeps increasing because of a memory leak in the kernel dentry >> cache. This is the same problem as outlined by others here: >> >> https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html >> >> So I was wondering whether this problem was fixed? I am using centos 6 with >> the following kernel: >> >> Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun 27 >> 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux >> >> At one point dentry was using 3GB plus on my 8GB system! >> >> I am currently using a cron job to clear the cache every so often: >> >> sync&& echo 2>/proc/sys/vm/drop_caches >> >> The above works but I am looking for a more permanent solution. To that end >> I tried increasing: >> >> echo 10000> /proc/sys/vm/vfs_cache_pressure >> >> And in /etc/sysctl.conf But to no effect. >> >> So any idea how to fix this? >> >> Regards, >> Abdussamad >> >> _______________________________________________ >> rhelv6-list mailing list >> rhelv6-list at redhat.com >> https://www.redhat.com/mailman/listinfo/rhelv6-list >> > > There are some things you can try to do. You can collect a vmcore from > a time period during which the system has exhausted nearly all of it's > memory due to this leak. Then you could try to analyze kmem to > indicate if there is a problem with the kernel. > > You could even go as simple as just looking at top, or the contents of > /proc//status periodically for the set of apps you suspect. If > you do have a single app that is leaking memory, you should be able to > record and graph a consistent increasing trend in the amount of memory > the faulty app is leaking. That would at least give you a starting > point for where to use an app like valgrind. > > Also, if you don't know which proc is at fault, I'd start with > /etc/crontab: > */5 * * * * root ps axo comm,vsize,rss | tail -n +2>> /tmp/rawdata > > This would collect /proc/pid/statm data for a while. > > This should work on an selinux-enforcing machine based on the output of: > # sesearch -As crond_t | grep tmp > > Then use awk to find the low- and high-water marks for each comm. > > You could get fancy and add timestamps to the records; maybe track > current, low-water, and high-water marks, then gnuplot with error > bars. > > As far as your drop_caches work around, know that drop_caches may > cause performance degrade because some cached data are flushed and > system have to load them from disk if they are needed again. > > Use the "ps aux" above results in the cron job and locate which > program have a growing RSS. > And sysstat (/var/log/sa/sar*) may provide some historical memory > information that you may be interested in. > > Hope this gives you somewhere to look. > > I will follow-up on the thread you mentioned. > > ~rp > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > I don't understand. Isn't dentry cache managed by the kernel? So why would I look at applications for possible leaks when its obviously the kernel that's at fault here? Please read my post again including the thread I linked to. It seems to me you've misunderstood my problem. From robinprice at gmail.com Mon Aug 29 23:56:18 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Mon, 29 Aug 2011 19:56:18 -0400 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <4E5C1051.7070207@abdussamad.com> Message-ID: On Mon, Aug 29, 2011 at 7:39 PM, robinprice at gmail.com wrote: > On Mon, Aug 29, 2011 at 6:18 PM, Abdussamad Abdurrazzaq > wrote: >> Hello >> >> Ok please ignore my previous email (if you've seen it). It's quite confused >> because I posted using gmane.org. >> >> I know about how Linux reports memory usage. My problem is very much real. >> Memory usage keeps increasing because of a memory leak in the kernel dentry >> cache. This is the same problem as outlined by others here: >> >> https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html >> >> So I was wondering whether this problem was fixed? I am using centos 6 with >> the following kernel: >> >> Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun 27 >> 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux >> >> At one point dentry was using 3GB plus on ?my 8GB system! >> >> I am currently using a cron job to clear the cache every so often: >> >> sync && echo 2 >/proc/sys/vm/drop_caches >> >> The above works but I am looking for a more permanent solution. To that end >> I tried increasing: >> >> echo 10000> /proc/sys/vm/vfs_cache_pressure >> >> And in /etc/sysctl.conf But to no effect. >> >> So any idea how to fix this? >> >> Regards, >> Abdussamad >> >> _______________________________________________ >> rhelv6-list mailing list >> rhelv6-list at redhat.com >> https://www.redhat.com/mailman/listinfo/rhelv6-list >> > > > There are some things you can try to do. You can collect a vmcore from > a time period during which the system has exhausted nearly all of it's > memory due to this leak. ?Then you could try to analyze kmem to > indicate if there is a problem with the kernel. > > You could even go as simple as just looking at top, or the contents of > /proc//status periodically for the set of apps you suspect. ?If > you do have a single app that is leaking memory, you should be able to > record and graph a consistent increasing trend in the amount of memory > the faulty app is leaking. ?That would at least give you a starting > point for where to use an app like valgrind. > > Also, if you don't know which proc is at fault, I'd start with > /etc/crontab: > ?*/5 * * * * root ps axo comm,vsize,rss | tail -n +2 >> /tmp/rawdata > > This would collect /proc/pid/statm data for a while. > > This should work on an selinux-enforcing machine based on the output of: > ?# sesearch -As crond_t | grep tmp > > Then use awk to find the low- and high-water marks for each comm. > > You could get fancy and add timestamps to the records; maybe track > current, low-water, and high-water marks, then gnuplot with error > bars. > > As far as your drop_caches work around, know that drop_caches may > cause performance degrade because some cached data are flushed and > system have to load them from disk if they are needed again. > > Use the "ps aux" above results in the cron job and locate which > program have a growing RSS. > And sysstat (/var/log/sa/sar*) may provide some historical memory > information that you may be interested in. > > Hope this gives you somewhere to look. > > I will follow-up on the thread you mentioned. > > ~rp > I would also encourage a lot of you to open a support case regarding this issue. But I know some of you are using other alternatives, so I will assist the best I can. I hope the latest --changelog might assist. I can't find anything related to said memory leak but will look more into it tomorrow. [rprice at x200 ~]$ uname -a > rhel6.1-changelog.txt [rprice at x200 ~]$ echo "---" >> rhel6.1-changelog.txt [rprice at x200 ~]$ rpm -qa kernel --changelog >> rhel6.1-changelog.txt http://people.redhat.com/rprice/rhel6.1-changelog.txt ~rp From abdussamad at abdussamad.com Tue Aug 30 00:14:40 2011 From: abdussamad at abdussamad.com (Abdussamad Abdurrazzaq) Date: Tue, 30 Aug 2011 05:14:40 +0500 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <4E5C1051.7070207@abdussamad.com> Message-ID: <4E5C2B70.2080406@abdussamad.com> On 08/30/2011 04:56 AM, robinprice at gmail.com wrote: > > > I would also encourage a lot of you to open a support case regarding > this issue. But I know some of you are using other alternatives, so I > will assist the best I can. I hope the latest --changelog might > assist. I can't find anything related to said memory leak but will > look more into it tomorrow. > > [rprice at x200 ~]$ uname -a> rhel6.1-changelog.txt > [rprice at x200 ~]$ echo "---">> rhel6.1-changelog.txt > [rprice at x200 ~]$ rpm -qa kernel --changelog>> rhel6.1-changelog.txt > > http://people.redhat.com/rprice/rhel6.1-changelog.txt > > ~rp > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > Well I am using a really old kernel so maybe I just have to wait for my distro to release the newer version. The one I am using is: 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun 27 And the latest one is 2.6.32-131.12.1 From robinprice at gmail.com Tue Aug 30 14:08:09 2011 From: robinprice at gmail.com (robinprice at gmail.com) Date: Tue, 30 Aug 2011 10:08:09 -0400 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5C2624.8070903@abdussamad.com> References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> Message-ID: I did some more searching this morning as I mentioned I would last night. I have not found anything in particular to your situation. The only suggestions I have would be: 1) Try getting a core during the memory consumption as I mentioned and do a RCA on the vmcore. 2) Write a stap script to trace d_alloc in the kernel (or one of the d_cache functions) to see who is allocating dentries, and correlate that to a process (a perf script would help you do that pretty easily). 3) Use lsof to see who has tons of open files. Presumably if you're swapping with 100% of ram holding dentries, someone is using those dentries which means lots of open files. Good luck. Sorry I couldn't find anything. If anyone has a valid RHEL subscription, I would encourage you to try with the latest RHEL6 kernel to see if the leak is still there, and if it is, allow GSS to help you find root cause. ~rp On Mon, Aug 29, 2011 at 7:52 PM, Abdussamad Abdurrazzaq wrote: > On 08/30/2011 04:39 AM, robinprice at gmail.com wrote: >> >> On Mon, Aug 29, 2011 at 6:18 PM, Abdussamad Abdurrazzaq >> ?wrote: >>> >>> Hello >>> >>> Ok please ignore my previous email (if you've seen it). It's quite >>> confused >>> because I posted using gmane.org. >>> >>> I know about how Linux reports memory usage. My problem is very much >>> real. >>> Memory usage keeps increasing because of a memory leak in the kernel >>> dentry >>> cache. This is the same problem as outlined by others here: >>> >>> https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html >>> >>> So I was wondering whether this problem was fixed? I am using centos 6 >>> with >>> the following kernel: >>> >>> Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun >>> 27 >>> 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux >>> >>> At one point dentry was using 3GB plus on ?my 8GB system! >>> >>> I am currently using a cron job to clear the cache every so often: >>> >>> sync&& ?echo 2>/proc/sys/vm/drop_caches >>> >>> The above works but I am looking for a more permanent solution. To that >>> end >>> I tried increasing: >>> >>> echo 10000> ?/proc/sys/vm/vfs_cache_pressure >>> >>> And in /etc/sysctl.conf But to no effect. >>> >>> So any idea how to fix this? >>> >>> Regards, >>> Abdussamad >>> >>> _______________________________________________ >>> rhelv6-list mailing list >>> rhelv6-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/rhelv6-list >>> >> >> There are some things you can try to do. You can collect a vmcore from >> a time period during which the system has exhausted nearly all of it's >> memory due to this leak. ?Then you could try to analyze kmem to >> indicate if there is a problem with the kernel. >> >> You could even go as simple as just looking at top, or the contents of >> /proc//status periodically for the set of apps you suspect. ?If >> you do have a single app that is leaking memory, you should be able to >> record and graph a consistent increasing trend in the amount of memory >> the faulty app is leaking. ?That would at least give you a starting >> point for where to use an app like valgrind. >> >> Also, if you don't know which proc is at fault, I'd start with >> /etc/crontab: >> ? */5 * * * * root ps axo comm,vsize,rss | tail -n +2>> ?/tmp/rawdata >> >> This would collect /proc/pid/statm data for a while. >> >> This should work on an selinux-enforcing machine based on the output of: >> ?# sesearch -As crond_t | grep tmp >> >> Then use awk to find the low- and high-water marks for each comm. >> >> You could get fancy and add timestamps to the records; maybe track >> current, low-water, and high-water marks, then gnuplot with error >> bars. >> >> As far as your drop_caches work around, know that drop_caches may >> cause performance degrade because some cached data are flushed and >> system have to load them from disk if they are needed again. >> >> Use the "ps aux" above results in the cron job and locate which >> program have a growing RSS. >> And sysstat (/var/log/sa/sar*) may provide some historical memory >> information that you may be interested in. >> >> Hope this gives you somewhere to look. >> >> I will follow-up on the thread you mentioned. >> >> ~rp >> >> _______________________________________________ >> rhelv6-list mailing list >> rhelv6-list at redhat.com >> https://www.redhat.com/mailman/listinfo/rhelv6-list >> > I don't understand. Isn't dentry cache managed by the kernel? So why would I > look at applications for possible leaks when its obviously the kernel that's > at fault here? Please read my post again including the thread I linked to. > ?It seems to me you've misunderstood my problem. > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > From wolfy at nobugconsulting.ro Tue Aug 30 20:13:54 2011 From: wolfy at nobugconsulting.ro (Manuel Wolfshant) Date: Tue, 30 Aug 2011 23:13:54 +0300 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> Message-ID: <4E5D4482.3060005@nobugconsulting.ro> On 08/30/2011 05:08 PM, robinprice at gmail.com wrote: > I did some more searching this morning as I mentioned I would last > night. I have not found anything in particular to your situation. > The only suggestions I have would be: > > 1) Try getting a core during the memory consumption as I mentioned and > do a RCA on the vmcore. > 2) Write a stap script to trace d_alloc in the kernel (or one of the > d_cache functions) to see who is allocating dentries, and correlate > that to a process (a perf script would help you do that pretty > easily). > 3) Use lsof to see who has tons of open files. Presumably if you're > swapping with 100% of ram holding dentries, someone is using those > dentries which means lots of open files. > > Good luck. Sorry I couldn't find anything. > > If anyone has a valid RHEL subscription, I would encourage you to try > with the latest RHEL6 kernel to see if the leak is still there, and if > it is, allow GSS to help you find root cause. > Or at least try the newer newer kernel from RHEL 6.1 as available in Scientific Linux: http://ftp.scientificlinux.org/linux/scientific/6.1/x86_64/updates/security/kernel-2.6.32-131.12.1.el6.x86_64.rpm > ~rp > > > On Mon, Aug 29, 2011 at 7:52 PM, Abdussamad Abdurrazzaq > wrote: >> On 08/30/2011 04:39 AM, robinprice at gmail.com wrote: >>> On Mon, Aug 29, 2011 at 6:18 PM, Abdussamad Abdurrazzaq >>> wrote: >>>> Hello >>>> >>>> Ok please ignore my previous email (if you've seen it). It's quite >>>> confused >>>> because I posted using gmane.org. >>>> >>>> I know about how Linux reports memory usage. My problem is very much >>>> real. >>>> Memory usage keeps increasing because of a memory leak in the kernel >>>> dentry >>>> cache. This is the same problem as outlined by others here: >>>> >>>> https://www.redhat.com/archives/rhelv6-list/2011-February/msg00001.html >>>> >>>> So I was wondering whether this problem was fixed? I am using centos 6 >>>> with >>>> the following kernel: >>>> >>>> Linux serve3.websitetheme.com. 2.6.32-71.29.1.el6.x86_64 #1 SMP Mon Jun >>>> 27 >>>> 19:49:27 BST 2011 x86_64 x86_64 x86_64 GNU/Linux >>>> >>>> At one point dentry was using 3GB plus on my 8GB system! >>>> >>>> I am currently using a cron job to clear the cache every so often: >>>> >>>> sync&& echo 2>/proc/sys/vm/drop_caches >>>> >>>> The above works but I am looking for a more permanent solution. To that >>>> end >>>> I tried increasing: >>>> >>>> echo 10000> /proc/sys/vm/vfs_cache_pressure >>>> >>>> And in /etc/sysctl.conf But to no effect. >>>> >>>> So any idea how to fix this? >>>> >>>> Regards, >>>> Abdussamad >>>> >>>> _______________________________________________ >>>> rhelv6-list mailing list >>>> rhelv6-list at redhat.com >>>> https://www.redhat.com/mailman/listinfo/rhelv6-list >>>> >>> There are some things you can try to do. You can collect a vmcore from >>> a time period during which the system has exhausted nearly all of it's >>> memory due to this leak. Then you could try to analyze kmem to >>> indicate if there is a problem with the kernel. >>> >>> You could even go as simple as just looking at top, or the contents of >>> /proc//status periodically for the set of apps you suspect. If >>> you do have a single app that is leaking memory, you should be able to >>> record and graph a consistent increasing trend in the amount of memory >>> the faulty app is leaking. That would at least give you a starting >>> point for where to use an app like valgrind. >>> >>> Also, if you don't know which proc is at fault, I'd start with >>> /etc/crontab: >>> */5 * * * * root ps axo comm,vsize,rss | tail -n +2>> /tmp/rawdata >>> >>> This would collect /proc/pid/statm data for a while. >>> >>> This should work on an selinux-enforcing machine based on the output of: >>> # sesearch -As crond_t | grep tmp >>> >>> Then use awk to find the low- and high-water marks for each comm. >>> >>> You could get fancy and add timestamps to the records; maybe track >>> current, low-water, and high-water marks, then gnuplot with error >>> bars. >>> >>> As far as your drop_caches work around, know that drop_caches may >>> cause performance degrade because some cached data are flushed and >>> system have to load them from disk if they are needed again. >>> >>> Use the "ps aux" above results in the cron job and locate which >>> program have a growing RSS. >>> And sysstat (/var/log/sa/sar*) may provide some historical memory >>> information that you may be interested in. >>> >>> Hope this gives you somewhere to look. >>> >>> I will follow-up on the thread you mentioned. >>> >>> ~rp >>> >>> _______________________________________________ >>> rhelv6-list mailing list >>> rhelv6-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/rhelv6-list >>> >> I don't understand. Isn't dentry cache managed by the kernel? So why would I >> look at applications for possible leaks when its obviously the kernel that's >> at fault here? Please read my post again including the thread I linked to. >> It seems to me you've misunderstood my problem. >> >> _______________________________________________ >> 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 abdussamad at abdussamad.com Tue Aug 30 20:34:10 2011 From: abdussamad at abdussamad.com (Abdussamad Abdurrazzaq) Date: Wed, 31 Aug 2011 01:34:10 +0500 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5D4482.3060005@nobugconsulting.ro> References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> <4E5D4482.3060005@nobugconsulting.ro> Message-ID: <4E5D4942.6010907@abdussamad.com> As I said before I'll just wait for my distro to release an updated kernel. CentOS is supposed to release the 6.1 version in a week or two according to what I read on their forum. From wolfy at nobugconsulting.ro Tue Aug 30 21:03:02 2011 From: wolfy at nobugconsulting.ro (Manuel Wolfshant) Date: Wed, 31 Aug 2011 00:03:02 +0300 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5D4942.6010907@abdussamad.com> References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> <4E5D4482.3060005@nobugconsulting.ro> <4E5D4942.6010907@abdussamad.com> Message-ID: <4E5D5006.3070903@nobugconsulting.ro> On 08/30/2011 11:34 PM, Abdussamad Abdurrazzaq wrote: > As I said before I'll just wait for my distro to release an updated > kernel. CentOS is supposed to release the 6.1 version in a week or two > according to what I read on their forum. you can wait if you so please. but do not bet any money on that release date. Centos 6.1 is not in QA yet. From abdussamad at abdussamad.com Tue Aug 30 21:11:54 2011 From: abdussamad at abdussamad.com (Abdussamad Abdurrazzaq) Date: Wed, 31 Aug 2011 02:11:54 +0500 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5D5006.3070903@nobugconsulting.ro> References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> <4E5D4482.3060005@nobugconsulting.ro> <4E5D4942.6010907@abdussamad.com> <4E5D5006.3070903@nobugconsulting.ro> Message-ID: <4E5D521A.80400@abdussamad.com> On 08/31/2011 02:03 AM, Manuel Wolfshant wrote: > On 08/30/2011 11:34 PM, Abdussamad Abdurrazzaq wrote: >> As I said before I'll just wait for my distro to release an updated >> kernel. CentOS is supposed to release the 6.1 version in a week or >> two according to what I read on their forum. > you can wait if you so please. but do not bet any money on that > release date. Centos 6.1 is not in QA yet. > > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list > It isn't? I thought it was. Anyway I can't go around experiment with different kernels on my production server. The clear cache via cron job solution is enough for now. From citros.airv at gmail.com Tue Aug 30 23:04:42 2011 From: citros.airv at gmail.com (Citros Airv) Date: Tue, 30 Aug 2011 16:04:42 -0700 Subject: [rhelv6-list] Tool to automate creation of install CDs. Message-ID: Hi List, I'm trying to put together a process to automate the creation of an installation CD for an appliance we manage. The goal is that given a set of RPMs needed on the target appliance we put together a custom CD containing only those RPMs. I was thinking of using the livecd-tools package but the resulting CD can't be used to install on a target system according to man page. Any pointers appreciated. --Citros -------------- next part -------------- An HTML attachment was scrubbed... URL: From RJM002 at SHSU.EDU Tue Aug 30 23:17:28 2011 From: RJM002 at SHSU.EDU (Marti, Robert) Date: Tue, 30 Aug 2011 18:17:28 -0500 Subject: [rhelv6-list] Tool to automate creation of install CDs. In-Reply-To: References: Message-ID: <68BC7805-09AD-4B17-8AD9-E7D91D8A6681@shsu.edu> Look into kickstarts. That's exactly what they're for. On Aug 30, 2011, at 18:07, "Citros Airv" wrote: > Hi List, > > I'm trying to put together a process to automate the creation of an installation CD for an appliance we manage. The goal is that given a set of RPMs needed on the target appliance we put together a custom CD containing only those RPMs. I was thinking of using the livecd-tools package but the resulting CD can't be used to install on a target system according to man page. > > Any pointers appreciated. > > > --Citros > _______________________________________________ > rhelv6-list mailing list > rhelv6-list at redhat.com > https://www.redhat.com/mailman/listinfo/rhelv6-list From citros.airv at gmail.com Tue Aug 30 23:33:33 2011 From: citros.airv at gmail.com (Citros Airv) Date: Tue, 30 Aug 2011 16:33:33 -0700 Subject: [rhelv6-list] Tool to automate creation of install CDs. In-Reply-To: <68BC7805-09AD-4B17-8AD9-E7D91D8A6681@shsu.edu> References: <68BC7805-09AD-4B17-8AD9-E7D91D8A6681@shsu.edu> Message-ID: Yes, thanks. I'm planing to use a kickstart file but my question is prior to that. The goal is to minimize the boot cd so that is not larger than 700MB. So given a set of RPMs I want to create the boot disk itself based on that...After that I'll use kickstart to drive the installation as suggested. --Citros On Tue, Aug 30, 2011 at 4:17 PM, Marti, Robert wrote: > Look into kickstarts. That's exactly what they're for. > > On Aug 30, 2011, at 18:07, "Citros Airv" wrote: > > > Hi List, > > > > I'm trying to put together a process to automate the creation of an > installation CD for an appliance we manage. The goal is that given a set of > RPMs needed on the target appliance we put together a custom CD containing > only those RPMs. I was thinking of using the livecd-tools package but the > resulting CD can't be used to install on a target system according to man > page. > > > > Any pointers appreciated. > > > > > > --Citros > > _______________________________________________ > > 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 paul.krizak at amd.com Wed Aug 31 00:16:54 2011 From: paul.krizak at amd.com (Paul Krizak) Date: Tue, 30 Aug 2011 17:16:54 -0700 Subject: [rhelv6-list] Tool to automate creation of install CDs. In-Reply-To: References: <68BC7805-09AD-4B17-8AD9-E7D91D8A6681@shsu.edu> Message-ID: <4E5D7D76.3030905@amd.com> The RHEL distro comes with a boot.iso that is a minimal environment that will get you into kickstart. If this is insufficient you can unpack the ISO and put whatever you want into it, rebuild the ISO and there you go. I'm assuming of course that the reason you want to do something like this is to include extra software so that during %pre install you can do things like partition your disks or query a DB or something? Though when we need to do this, we simply put the tools we need on an NFS filesystem and mount it from %pre and fetch the tools from there. That way there's no need to rebuild ISOs and stuff. Remember that during %post install you've already got a fully installed system at /mnt/local, so just make sure whatever packages you need get installed during %packages phase. Paul Krizak 7171 Southwest Pkwy MS B200.3A MTS Systems Engineer Austin, TX 78735 Advanced Micro Devices Desk: (512) 602-8775 Linux/Unix Systems Engineering Cell: (512) 791-0686 Global IT Infrastructure Fax: (512) 602-0468 On 08/30/2011 04:33 PM, Citros Airv wrote: > Yes, thanks. I'm planing to use a kickstart file but my question is > prior to that. The goal is to minimize the boot cd so that is not larger > than 700MB. So given a set of RPMs I want to create the boot disk itself > based on that...After that I'll use kickstart to drive the installation > as suggested. > > --Citros > > On Tue, Aug 30, 2011 at 4:17 PM, Marti, Robert > wrote: > > Look into kickstarts. That's exactly what they're for. > > On Aug 30, 2011, at 18:07, "Citros Airv" > wrote: > > > Hi List, > > > > I'm trying to put together a process to automate the creation of > an installation CD for an appliance we manage. The goal is that > given a set of RPMs needed on the target appliance we put together a > custom CD containing only those RPMs. I was thinking of using the > livecd-tools package but the resulting CD can't be used to install > on a target system according to man page. > > > > Any pointers appreciated. > > > > > > --Citros > > _______________________________________________ > > 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 mlist at lubrical.net Wed Aug 31 06:37:48 2011 From: mlist at lubrical.net (Tim) Date: Wed, 31 Aug 2011 16:37:48 +1000 Subject: [rhelv6-list] Tool to automate creation of install CDs. In-Reply-To: References: <68BC7805-09AD-4B17-8AD9-E7D91D8A6681@shsu.edu> Message-ID: > Yes, thanks. I'm planing to use a kickstart file but my question is prior > to > that. The goal is to minimize the boot cd so that is not larger than > 700MB. > So given a set of RPMs I want to create the boot disk itself based on > that...After that I'll use kickstart to drive the installation as > suggested. You can efficiently kickstart from the CD/DVD you boot from. Build an iso that has a ks.cfg on it and set up grub to pass it automatically, like "ks=cdrom:/ks.cfg". This will make a self contained CD that will kickstart from itself when booted. -- Tim From john.horne at plymouth.ac.uk Wed Aug 31 12:03:34 2011 From: john.horne at plymouth.ac.uk (John Horne) Date: Wed, 31 Aug 2011 13:03:34 +0100 Subject: [rhelv6-list] Setting default language Message-ID: <1314792214.1644.9.camel@jhorne.csd.plymouth.ac.uk> Hello, Running RHEL 6.1, I have set the default language in /etc/sysconfig/i18n as LANG="en_GB.UTF-8". However, when I log in as a user I get a different language: echo $LANG en_US.UTF-8 If I then 'su -' to root, and do the same then I get the correct language: echo $LANG en_GB.UTF-8 Any ideas why the non-root user does not get the correct language? The 'English (UK)' language was selected during installation. I have also tried running 'system-config-language', but the problem persists. Thanks, John. -- John Horne Tel: +44 (0)1752 587287 Plymouth University, UK Fax: +44 (0)1752 587001 From gianluca.cecchi at gmail.com Wed Aug 31 12:36:59 2011 From: gianluca.cecchi at gmail.com (Gianluca Cecchi) Date: Wed, 31 Aug 2011 14:36:59 +0200 Subject: [rhelv6-list] Setting default language In-Reply-To: <1314792214.1644.9.camel@jhorne.csd.plymouth.ac.uk> References: <1314792214.1644.9.camel@jhorne.csd.plymouth.ac.uk> Message-ID: On Wed, Aug 31, 2011 at 2:03 PM, John Horne wrote: > Any ideas why the non-root user does not get the correct language? > The 'English (UK)' language was selected during installation. > I have also tried running 'system-config-language', but the problem > persists. Just a guess, after watching @ one rh el 6.1 system I have. I notice that under /etc/profile.d directory there is a "lang.sh" script that checks for a file named ".i18n" under the home directory of the user that logs in. The files under this directory are executed by /etc/profile at login time. Is there any .i18n file inside the home directory of this user? In that case it seems it is parsed and $LANG set accordingly.... HIH, Gianluca From john.horne at plymouth.ac.uk Wed Aug 31 12:51:27 2011 From: john.horne at plymouth.ac.uk (John Horne) Date: Wed, 31 Aug 2011 13:51:27 +0100 Subject: [rhelv6-list] Setting default language In-Reply-To: References: <1314792214.1644.9.camel@jhorne.csd.plymouth.ac.uk> Message-ID: <1314795087.1644.10.camel@jhorne.csd.plymouth.ac.uk> On Wed, 2011-08-31 at 14:36 +0200, Gianluca Cecchi wrote: > On Wed, Aug 31, 2011 at 2:03 PM, John Horne wrote: > > > Any ideas why the non-root user does not get the correct language? > > The 'English (UK)' language was selected during installation. > > I have also tried running 'system-config-language', but the problem > > persists. > > Just a guess, after watching @ one rh el 6.1 system I have. > I notice that under /etc/profile.d directory there is a "lang.sh" > script that checks for a file named ".i18n" under the home directory > of the user that logs in. > > The files under this directory are executed by /etc/profile at login time. > > Is there any .i18n file inside the home directory of this user? > No, there is no .i18n file in the users home directory. John. -- John Horne Tel: +44 (0)1752 587287 Plymouth University, UK Fax: +44 (0)1752 587001 From kmazurek at neotek.waw.pl Wed Aug 31 15:21:41 2011 From: kmazurek at neotek.waw.pl (Krzysztof Mazurek) Date: Wed, 31 Aug 2011 17:21:41 +0200 Subject: [rhelv6-list] 10GbE devices under RHEL 5/6 Message-ID: Hi :) I looking at the 10Gbit Ethernet Adapters for "IBM Blades". I noticed the Emulex 10GbE Virtual Fabric Adapter II Advanced - a new addition to the offer. Does any one know, heaving such a card in the system (assuming it has a FCoE support) will I be seen as both - a FC card and Ethernet (1Gbit/s) ? I'm aware of limitations of blade servers and I want to use 10GbE pass thru from chassis to FCoE enabled 10GbE independent switch in RACK. What I don't want is next addon cards and switcher for ethernet (TCP/IP) communications ... We plan to implement quite a large EnterprsieDB database instance on that machines ;) Krzysztof -------------- next part -------------- An HTML attachment was scrubbed... URL: From solarflow99 at gmail.com Wed Aug 31 21:20:46 2011 From: solarflow99 at gmail.com (solarflow99) Date: Wed, 31 Aug 2011 17:20:46 -0400 Subject: [rhelv6-list] Kernel memory leak? In-Reply-To: <4E5D521A.80400@abdussamad.com> References: <4E5C1051.7070207@abdussamad.com> <4E5C2624.8070903@abdussamad.com> <4E5D4482.3060005@nobugconsulting.ro> <4E5D4942.6010907@abdussamad.com> <4E5D5006.3070903@nobugconsulting.ro> <4E5D521A.80400@abdussamad.com> Message-ID: Could you just use RHEL for your production server? Its free, just unsupported unless you want a paid support plan. On Tue, Aug 30, 2011 at 5:11 PM, Abdussamad Abdurrazzaq < abdussamad at abdussamad.com> wrote: > On 08/31/2011 02:03 AM, Manuel Wolfshant wrote: > >> On 08/30/2011 11:34 PM, Abdussamad Abdurrazzaq wrote: >> >>> As I said before I'll just wait for my distro to release an updated >>> kernel. CentOS is supposed to release the 6.1 version in a week or two >>> according to what I read on their forum. >>> >> you can wait if you so please. but do not bet any money on that release >> date. Centos 6.1 is not in QA yet. >> >> ______________________________**_________________ >> rhelv6-list mailing list >> rhelv6-list at redhat.com >> https://www.redhat.com/**mailman/listinfo/rhelv6-list >> >> It isn't? I thought it was. Anyway I can't go around experiment with > different kernels on my production server. The clear cache via cron job > solution is enough for now. > > > ______________________________**_________________ > 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: