Re: SLES 9 coming up with / read-only - SOLVED
Mark Post wrote: On 8/6/2008 at 3:51 PM, in message <[EMAIL PROTECTED]>, Matt Gourley <[EMAIL PROTECTED]> wrote: -snip- I've already tried running an fsck on the filesystem. Because I can't write to it, I wound up shutting down and logging off the problem system, LINKing its 201 disk MR to another running Linux system via hcp(1), setting it online, and running fsck on it. No problems found. Did you do an fsck, or an "fsck -f" command? What specific kernel level are you at? I ran an "fsck -f /dev/dasdb1", shutdown, came back up and all is well. Thanks to Mark and Richard for giving me a couple more tools for the toolbox. -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: SLES 9 coming up with / read-only
Mark Post wrote: On 8/6/2008 at 3:51 PM, in message <[EMAIL PROTECTED]>, Matt Gourley <[EMAIL PROTECTED]> wrote: -snip- I've already tried running an fsck on the filesystem. Because I can't write to it, I wound up shutting down and logging off the problem system, LINKing its 201 disk MR to another running Linux system via hcp(1), setting it online, and running fsck on it. No problems found. Did you do an fsck, or an "fsck -f" command? What specific kernel level are you at? Mark Post I did an "fsck" only. I did not run "fsck -f" Our kernel is 2.6.5-7.308-s390x. Also, I was able to at least remount the device r/w using Richard Gasiorowski's suggestion of "mount -t ext2 -o rw,remount /dev/dasdb1 /" When I do a "runlevel" it's telling me I'm in single-user mode. I'm going to try to specify runlevel 3, though that's listed as the default in /etc/inittab. -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: SLES 9 coming up with / read-only
Mark Post wrote: On 8/6/2008 at 3:28 PM, in message <[EMAIL PROTECTED]>, Matt Gourley <[EMAIL PROTECTED]> wrote: One of our Linux images was booted on Monday for the first time since March. During the boot, we started getting error messages telling us that / was a read-only file system: Is this repeatable, i.e., if you try to reboot does it do the same thing? If so, I would try booting into the installation system, and see if I could do an fsck on the file system. Normally this sort of thing happens when there's some sort of I/O error on the disk, but there have been some bugs about this happening on recoverable errors that don't actually make it to the console, since they were recovered. What is your maintenance level on that system? This is definitely repeatable. Every time we boot this system it comes up with the / filesystem as read-only. I've already tried running an fsck on the filesystem. Because I can't write to it, I wound up shutting down and logging off the problem system, LINKing its 201 disk MR to another running Linux system via hcp(1), setting it online, and running fsck on it. No problems found. I can still login to the system as root, though: # cat /etc/SuSE-release SUSE LINUX Enterprise Server 9 (s390x) VERSION = 9 PATCHLEVEL = 4 -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
SLES 9 coming up with / read-only
One of our Linux images was booted on Monday for the first time since March. During the boot, we started getting error messages telling us that / was a read-only file system: Waiting for device /dev/dasdb1 to appear: ok rootfs: major=94 minor=5 devn=24069 rootfs: /sys/block/dasdb/dasdb1 major=94 minor=5 devn=24069 umount: /sys: device is busy umount: /sys: device is busy VFS: Mounted root (ext2 filesystem) readonly. Trying to move old root to /initrd ... /initrd does not exist. Ignored. Unmounting old root Trying to free ramdisk memory ... failed Freeing unused kernel memory: 116k freed INIT: version 2.85 booting ... device-mapper: Allocated new minor_bits array for 1024 devices device-mapper: 4.4.0-ioctl (2005-01-12) initialised: [EMAIL PROTECTED] Activating device mapper... rm: cannot remove `/dev/mapper/control': Read-only file system ..failed ... I've double-checked the 201 disk, and it's set to be R/W: DASD 0201 3390 V6SHR1 R/W 3338 CYL ON DASD 4217 SUBCHANNEL = 0003 Nothing else is accessing this disk. Apparently, SLES thinks / needs to be mounted readonly. Any ideas? Thanks, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: BaseVol/GuestVol server for SLES9
(Sorry this is so late; we wanted to make sure we had something solid before I responded.) Michael MacIsaac wrote: What are other people using for the layout of their master/golden image? Right now, we're testing out the basevol/guestvol setup at http://www.linuxvm.org/Info/HOWTOs/basevol9.html With the help of the folks at Novell, we finally solved our FCP issue, allowing our guestvol systems to attach to FCP LUNs for filesystems like /opt, and save DASD for the R/W portions of the operating system. On our basevol system, we have a file called /etc/luns.guestvols, which looks like this: LINUX15:0.0.f106:0x50050768012004cc:0x0006 LINUX1:0.0.f105:0x50050768012004cc:0x0003 LINUX5:0.0.f100:0x5005076300c99168:0x5702 LINUX16:0.0.f107:0x50050768012004cc:0x0007 LINUX9:0.0.f101:0x50050768012004cc:0x0001 LINUX9:0.0.f103:0x50050768012004cc:0x0005 The fields, separated by colons, are Guest ID, channel, WWPN, and LUN. We also have a script called /etc/init.d/boot.fcpluns: #!/bin/sh # Get the name of the guest: modprobe vmcp sleep 1 GUESTID=$(vmcp q userid | awk '{ print $1 }') # Get a list of the targets FCPTARGET=$(sed "/^$GUESTID/p;d" /etc/luns.guestvols) # iterate for a in $FCPTARGET ; do CCW=$(echo $a | awk -F":" '{ print $2 }') WWPN=$(echo $a | awk -F":" '{ print $3 }') LUN=$(echo $a | awk -F":" '{ print $4 }') /sbin/zfcp_host_configure $CCW 1 /sbin/zfcp_disk_configure $CCW $WWPN $LUN 1 done /etc/init.d/boot.fcpluns is run right after /etc/init.d/boot.guestvol during the boot process. When we add a system with an FCP LUN to our VM, we modify /etc/luns.guestvols and re-run /etc/init.d/make-guestvol to ensure the latest changes to this file will be sent to the new system. -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Basevol/Guestvol and ZFCP
Mark Post wrote: On Fri, May 4, 2007 at 2:00 PM, in message <[EMAIL PROTECTED]>, Matt Gourley <[EMAIL PROTECTED]> wrote: -snip- 2) If I add the LUN to the guestvol system through YaST and reboot, the LUN information is lost. Obviously, this isn't good. :) -snip- Sounds like you need to re-run mkinitrd and zipl. There's where I'm running into issues. I can't run mkinitrd and zipl from the guestvol system, as it's seeing /sys and /boot as read-only. I have added the following line to /etc/sysconfig/kernel on the basevol: INITRD_MODULES="dasd_fba_mod zfcp" and have run mkinitrd and zipl on the basevol. Without this, I would not have been able to see the LUN. When we setup the DASD beyond the 202 disk on the other guests, we had a similar problem. The solution was to write a script to mount any ext[23] filesystem in /etc/fstab that wasn't already mounted or /guestvol, and put that in /etc/init.d/rc3.d between the system scripts and the application scripts. -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Basevol/Guestvol and ZFCP
Hello, all. We're working on trying to save DASD by migrating a number of our systems over to basevol/guestvol , per the HowTo at http://linuxvm.org/Info/HOWTOs/basevol9.html . We have already migrated three of our Linux guests over to be guestvol systems; the migration was fairly painless and the extra mini-disk DASD (203 seen as /dev/dasdd1 mounted to /wasprofiles, for example) attached without a major glitch. We have a few systems that are using SVC LUNs that attach to Linux through ZFCP, however, and we're running into a couple of issues: 1) I will add the LUN to the guestvol system through YaST with the same WWPN and LUN number as the "old" system, and get a smaller LUN than both SVC and the "old" system expect. i.e. I mount /dev/sda1 on the guestvol system and see a 5GB LUN, but SVC says it's a 6GB LUN, and it is if I boot to the "old" system and mount it there. 2) If I add the LUN to the guestvol system through YaST and reboot, the LUN information is lost. Obviously, this isn't good. :) Anyone out there doing both basevol/guestvol and ZFCP? Thanks in advance for your help. -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Backups with FDR/Upstream
An answer! I re-ran my restore job with one parm to UPSTREAM added: DISASTERRECOVERY Y This re-created all my /dev files. I tried to IPL from LINUX5, and got an error message that kicked me out to CP; I re-mounted LINUX5's 100 disk on linux7, chroot'd to it, and ran zipl. Unmounted, unlinked, and linux5 IPL'd just fine. Big thanks to all for your help. -Matt Adam Thornton wrote: On Mar 22, 2006, at 1:14 PM, Matt Gourley wrote: The plot thickens... I checked /mnt/linux5/var/log/messages, and saw a number of these entries: Mar 22 13:54:10 (none) mingetty[777]: /dev/ttyS0: No such file or directory Upon checking /mnt/linux5/dev, I found that its well below its complement of devices. I checked /mnt/linux5/proc, and it too seems to be lacking directory structures and files. Maybe it's a backup issue, maybe it's a restore issue, but it's certainly an Upstream issue. I will find out what's going on, and keep you all posted. /proc is going to be empty; all the stuff in there is dynamically created as a result of various device drivers and kernel functions. /dev, on the other hand Well, if you're running udev then a lot of the nodes will be dynamically created. On the other hand, it sure smells like it can't find a console. Are you sure FDR is set to back up device nodes as well as regular files? Adam -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Backups with FDR/Upstream
The plot thickens... I checked /mnt/linux5/var/log/messages, and saw a number of these entries: Mar 22 13:54:10 (none) mingetty[777]: /dev/ttyS0: No such file or directory Upon checking /mnt/linux5/dev, I found that its well below its complement of devices. I checked /mnt/linux5/proc, and it too seems to be lacking directory structures and files. Maybe it's a backup issue, maybe it's a restore issue, but it's certainly an Upstream issue. I will find out what's going on, and keep you all posted. Thanks, -Matt Marist EDU wrote: Maybe you need to re-write the boot record. With the linux05 / mounted to /mnt/linux5 try: chroot /mnt/linux5 zipl Just a guess. Josh Konkol, CCSE CNE MCSE Technical Research Specialist .~.GuideOne Insurance /V\ /( )\ ^^-^^ [EMAIL PROTECTED] -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Posted At: Wednesday, March 22, 2006 9:35 AM Posted To: Marist EDU Conversation: Backups with FDR/Upstream Subject: Backups with FDR/Upstream Morning all, I know this isn't the Upstream list, but I figured I'd sanity-check what I've done with my Linux images setup and describe what I've run into in the hope that someone else has been down this road. :) I've setup two virtual servers under z/VM: - linux7, which is setup similarly to linux00 under the LPAR2VS Redbook: Class BG, OPTION LNKNOPAS. I've also setup Hipersockets between it and z/OS for Upstream purposes, and backups/restores of files and directories have been tested without issues. - linux5, my guinea-pig image. With LINUX5 shutdown and logged off VM, I've run the following commands on linux7 (big thanks to Mike MacIsaac et. al. and their clone script from LPAR2VS): hcp LINK LINUX5 100 1100 MR echo 1 > /sys/bus/ccw/devices/0.0.1100/online mkdir /mnt/linux5 mount /dev/dasdd1 /mnt/linux5 # Note: I'd checked the device name beforehand Now I've got linux5's / disk mounted on linux7. I then run a backup job from z/OS that calls Upstream on linux7 to backup /mnt/linux5/*. Now that I have a (supposed) full backup of linux5, I'm going to delete everything from /mnt/linux5 to test a full-system restore: cd /mnt/linux5 rm -rf * I then run a restore job from z/OS that restores linux7:/mnt/linux5 using Upstream. Finally, I clean up my links to make sure LINUX5 has its disk all to itself: umount /mnt/linux5 echo 0 > /sys/bus/ccw/devices/0.0.1100/online hcp DETACH 1100 Now I should be able to logon to LINUX5 and IPL a restored copy of everything from its 100 disk. However, when I do so, I get about 1/3 of the way through the boot process, and it's hanging on: qdio: loading QDIO base support version 2 ($Revision: 1.79.2.8 $/$Revision: 1.57 $/$Revision: 1.23.2.2 $) qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.54 $/$Revision: 1 .98.2.27 $/$Revision: 1.27.2.9 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.3 $/$Revi sion: 1.5.2.6 $/$Revision: 1.19.2.16 $ :IPv6 :VLAN) There's nothing in the scroll of the IPL before this point that inidicates any problems, though if necessary I could copy/paste the whole thing. Any insights into this would be helpful. Thanks, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Backups with FDR/Upstream
I tried both this and running `sync;sync;sync;umount /mnt/linux5` as Mike MacIsaac suggested. No dice; I keep stopping at the same point in the boot. I'm going to re-mount /mnt/linux5 on linux7 to see if anything has been written in the logs. -Matt Marist EDU wrote: Maybe you need to re-write the boot record. With the linux05 / mounted to /mnt/linux5 try: chroot /mnt/linux5 zipl Just a guess. Josh Konkol, CCSE CNE MCSE Technical Research Specialist .~.GuideOne Insurance /V\ /( )\ ^^-^^ [EMAIL PROTECTED] -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Posted At: Wednesday, March 22, 2006 9:35 AM Posted To: Marist EDU Conversation: Backups with FDR/Upstream Subject: Backups with FDR/Upstream Morning all, I know this isn't the Upstream list, but I figured I'd sanity-check what I've done with my Linux images setup and describe what I've run into in the hope that someone else has been down this road. :) I've setup two virtual servers under z/VM: - linux7, which is setup similarly to linux00 under the LPAR2VS Redbook: Class BG, OPTION LNKNOPAS. I've also setup Hipersockets between it and z/OS for Upstream purposes, and backups/restores of files and directories have been tested without issues. - linux5, my guinea-pig image. With LINUX5 shutdown and logged off VM, I've run the following commands on linux7 (big thanks to Mike MacIsaac et. al. and their clone script from LPAR2VS): hcp LINK LINUX5 100 1100 MR echo 1 > /sys/bus/ccw/devices/0.0.1100/online mkdir /mnt/linux5 mount /dev/dasdd1 /mnt/linux5 # Note: I'd checked the device name beforehand Now I've got linux5's / disk mounted on linux7. I then run a backup job from z/OS that calls Upstream on linux7 to backup /mnt/linux5/*. Now that I have a (supposed) full backup of linux5, I'm going to delete everything from /mnt/linux5 to test a full-system restore: cd /mnt/linux5 rm -rf * I then run a restore job from z/OS that restores linux7:/mnt/linux5 using Upstream. Finally, I clean up my links to make sure LINUX5 has its disk all to itself: umount /mnt/linux5 echo 0 > /sys/bus/ccw/devices/0.0.1100/online hcp DETACH 1100 Now I should be able to logon to LINUX5 and IPL a restored copy of everything from its 100 disk. However, when I do so, I get about 1/3 of the way through the boot process, and it's hanging on: qdio: loading QDIO base support version 2 ($Revision: 1.79.2.8 $/$Revision: 1.57 $/$Revision: 1.23.2.2 $) qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.54 $/$Revision: 1 .98.2.27 $/$Revision: 1.27.2.9 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.3 $/$Revi sion: 1.5.2.6 $/$Revision: 1.19.2.16 $ :IPv6 :VLAN) There's nothing in the scroll of the IPL before this point that inidicates any problems, though if necessary I could copy/paste the whole thing. Any insights into this would be helpful. Thanks, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Backups with FDR/Upstream
Morning all, I know this isn't the Upstream list, but I figured I'd sanity-check what I've done with my Linux images setup and describe what I've run into in the hope that someone else has been down this road. :) I've setup two virtual servers under z/VM: - linux7, which is setup similarly to linux00 under the LPAR2VS Redbook: Class BG, OPTION LNKNOPAS. I've also setup Hipersockets between it and z/OS for Upstream purposes, and backups/restores of files and directories have been tested without issues. - linux5, my guinea-pig image. With LINUX5 shutdown and logged off VM, I've run the following commands on linux7 (big thanks to Mike MacIsaac et. al. and their clone script from LPAR2VS): hcp LINK LINUX5 100 1100 MR echo 1 > /sys/bus/ccw/devices/0.0.1100/online mkdir /mnt/linux5 mount /dev/dasdd1 /mnt/linux5 # Note: I'd checked the device name beforehand Now I've got linux5's / disk mounted on linux7. I then run a backup job from z/OS that calls Upstream on linux7 to backup /mnt/linux5/*. Now that I have a (supposed) full backup of linux5, I'm going to delete everything from /mnt/linux5 to test a full-system restore: cd /mnt/linux5 rm -rf * I then run a restore job from z/OS that restores linux7:/mnt/linux5 using Upstream. Finally, I clean up my links to make sure LINUX5 has its disk all to itself: umount /mnt/linux5 echo 0 > /sys/bus/ccw/devices/0.0.1100/online hcp DETACH 1100 Now I should be able to logon to LINUX5 and IPL a restored copy of everything from its 100 disk. However, when I do so, I get about 1/3 of the way through the boot process, and it's hanging on: qdio: loading QDIO base support version 2 ($Revision: 1.79.2.8 $/$Revision: 1.57 $/$Revision: 1.23.2.2 $) qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.54 $/$Revision: 1 .98.2.27 $/$Revision: 1.27.2.9 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.3 $/$Revi sion: 1.5.2.6 $/$Revision: 1.19.2.16 $ :IPv6 :VLAN) There's nothing in the scroll of the IPL before this point that inidicates any problems, though if necessary I could copy/paste the whole thing. Any insights into this would be helpful. Thanks, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: zfcp and SLES 9 64-bit
This nailed it. I ran the mkinitrd command tailored to my environment, ran zipl, and IPL'd. /dev/sda mounted on boot. Big thanks. -Matt P David Schaub wrote: Matt Gourley wrote: I've run into an interesting issue with SLES 9 64-bit and zfcp and was hoping someone here had an idea to fix it. I've checked the archives but saw nothing matching this. I've got a 50G FCP LUN on device F100 attached to a VM named LINUX2. If I IPL SLES on LINUX2, zfcp discovers 0xF100 after /etc/init.d runs through all its startup scripts. I then need to log in as root to manually mount /dev/sda1 to my mountpoint, which happens with no trouble. Since this is a test environment, this is not a big deal, but when we move into production, this kind of intervention is less than optimal. (Call me crazy, but I'd like to avoid the 3am phone call to mount a filesystem that should be mounted on boot. :) ) Any ideas? You might try a mkinitrd to refresh the /boot/initrd file. I had an issue a few weeks back that DASD was added to an image but not recognized until late in the bootup process. Running: cd /boot mkinitrd -k vmlinux-2.6.5-7.244-s390x.gz -i initrd-2.6.5-7.244-s390x zipl may fix the issue. Make sure the DASD that you are expecting is accurately identifed. Also note that your version may differ from mine. David -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. If the reader of this message is not the intended recipient, you are hereby notified that your access is unauthorized, and any review, dissemination, distribution or copying of this message including any attachments is strictly prohibited. If you are not the intended recipient, please contact the sender and delete the material from any computer. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: zfcp and SLES 9 64-bit
Yeah. Probably should have mentioned that. :) Thanks, -Matt Neale Ferguson wrote: Silly question but your /etc/fstab has an entry for that device? -Original Message- I've got a 50G FCP LUN on device F100 attached to a VM named LINUX2. If I IPL SLES on LINUX2, zfcp discovers 0xF100 after /etc/init.d runs through all its startup scripts. I then need to log in as root to manually mount /dev/sda1 to my mountpoint, which happens with no trouble. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
zfcp and SLES 9 64-bit
Hi, I've run into an interesting issue with SLES 9 64-bit and zfcp and was hoping someone here had an idea to fix it. I've checked the archives but saw nothing matching this. I've got a 50G FCP LUN on device F100 attached to a VM named LINUX2. If I IPL SLES on LINUX2, zfcp discovers 0xF100 after /etc/init.d runs through all its startup scripts. I then need to log in as root to manually mount /dev/sda1 to my mountpoint, which happens with no trouble. Since this is a test environment, this is not a big deal, but when we move into production, this kind of intervention is less than optimal. (Call me crazy, but I'd like to avoid the 3am phone call to mount a filesystem that should be mounted on boot. :) ) Any ideas? Thanks, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
s390 and s390x directories in /nfs/sles9root/core9
All, Due to vendor issues, we here have been moving back and forth between 31-bit and 64-bit SLES9, using the LPAR to Virtual Servers Redbook as our guide. We're settled (for now) on 64-bit, after being told that this will never, ever change, that the ports and migrations we're going to move to will be on 64-bit SLES9. Of course, we're a little skeptical, so when we moved everything over to 64-bit, we kept the 31-bit rpms on LINUX00. So our directory structure looks like: /nfs/sles9root/core9/CD?/suse/s390 and /nfs/sles9root/core9/CD?/suse/s390x When I try to, for example, run 'yast -i XFree86-libs-32bit' on one of our clones, yast exits with no errors, but XFree86-libs-32bit is not installed either. Is keeping the 31-bit code an issue? Thanks in advance, -Matt -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
Mark, Thanks for your help on this. Unfortunately, we had to punt on this and do a re-install of the golden image. It was either that or tell management we still don't have the thing setup the way we want in next week's staff meeting. ;-) Maybe, once we get settled and are happily cloning, I'll create a new "golden" image and break it the same way, just to find out what I screwed up. -Matt Post, Mark K wrote: Matt, Ok, this is actually good news. The changes between 2.4 and 2.6 in how devices get initialized are massive, but the fact that both the modules load and don't generate errors means that you are close. Unfortunately, I haven't had the time to play with 2.6 stuff, so I don't think I can take you any further, other than to say that an "ifup" command should work now. If your network was configured previously, it should still be so. Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Monday, June 27, 2005 12:22 PM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation lsmod does confirm that it's loaded. ifconfig -a, on the other hand, still doesn't list anything. If I try to manually bring the interface up by 'ifup qeth-bus-ccw-0.0.0700' I get "Interface is not available." -Matt Post, Mark K wrote: Well, that looks like it worked. lsmod will confirm if it loaded, and ifconfig -a should show an interface now (maybe). Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Monday, June 27, 2005 11:58 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation # modprobe qeth qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.20 $/$Revision: 1 .98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.1 $/$Revi sion: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VLAN) Jun 27 11:56:35 LNXMSTR9 kernel: qeth: loading qeth S/390 OSA-Express driver ($R evision: 1.77.2.20 $/$Revision: 1.98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8 .2.2 $/$Revision: 1.7.2.1 $/$Revision: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VL AN) -Matt Post, Mark K wrote: Ok, then what happens if you do a "modprobe qeth" command? Mark Post -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
lsmod does confirm that it's loaded. ifconfig -a, on the other hand, still doesn't list anything. If I try to manually bring the interface up by 'ifup qeth-bus-ccw-0.0.0700' I get "Interface is not available." -Matt Post, Mark K wrote: Well, that looks like it worked. lsmod will confirm if it loaded, and ifconfig -a should show an interface now (maybe). Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Monday, June 27, 2005 11:58 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation # modprobe qeth qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.20 $/$Revision: 1 .98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.1 $/$Revi sion: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VLAN) Jun 27 11:56:35 LNXMSTR9 kernel: qeth: loading qeth S/390 OSA-Express driver ($R evision: 1.77.2.20 $/$Revision: 1.98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8 .2.2 $/$Revision: 1.7.2.1 $/$Revision: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VL AN) -Matt Post, Mark K wrote: Ok, then what happens if you do a "modprobe qeth" command? Mark Post -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
# modprobe qeth qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.77.2.20 $/$Revision: 1 .98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8.2.2 $/$Revision: 1.7.2.1 $/$Revi sion: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VLAN) Jun 27 11:56:35 LNXMSTR9 kernel: qeth: loading qeth S/390 OSA-Express driver ($R evision: 1.77.2.20 $/$Revision: 1.98.2.11 $/$Revision: 1.27.2.5 $/$Revision: 1.8 .2.2 $/$Revision: 1.7.2.1 $/$Revision: 1.5.2.4 $/$Revision: 1.19.2.7 $ :IPv6 :VL AN) -Matt Post, Mark K wrote: Ok, then what happens if you do a "modprobe qeth" command? Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Monday, June 27, 2005 8:39 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation 'lsmod | grep qdio' comes up with information; 'lsmod | grep qeth' does not. I think this is where we're running into issues, as an interface does not get created. 'dmesg' doesn't say anything about either qdio or qeth. Thanks, -Matt Post, Mark K wrote: If you want to avoid re-installing, we'll need diagnostic information. Do the qeth and qdio kernel modules load successfully? (lsmod) Do they create an interface? (ifconfig -a) If not, what shows up in the kernel ring buffer? (dmesg) Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Friday, June 24, 2005 9:10 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation Well, I did spend some time getting rid of a bunch of RPMs on the golden image yesterday. Unfortunately for me, one or more of the RPMs I nuked were necessary for the network configuration (qeth). The golden image boots, but cannot connect to the network. Luckily, we had already setup our controller per the book, and I can copy any RPM from the source tree (/nfs) to the golden image (/sles9golden), reboot the golden image, and re-install the rpm. I've already done this for the vlan RPM, but the golden image still will not come up on the network. Short of listing every RPM I purged, can anyone out there tell me what RPMs are necessary for qeth to run? Or is there further configuration I need to do on my golden image? (Oh boy, a chance to work on my ed(1) skills! :-) ) -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
'lsmod | grep qdio' comes up with information; 'lsmod | grep qeth' does not. I think this is where we're running into issues, as an interface does not get created. 'dmesg' doesn't say anything about either qdio or qeth. Thanks, -Matt Post, Mark K wrote: If you want to avoid re-installing, we'll need diagnostic information. Do the qeth and qdio kernel modules load successfully? (lsmod) Do they create an interface? (ifconfig -a) If not, what shows up in the kernel ring buffer? (dmesg) Mark Post -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Matt Gourley Sent: Friday, June 24, 2005 9:10 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: RPMs to keep/purge after installation Well, I did spend some time getting rid of a bunch of RPMs on the golden image yesterday. Unfortunately for me, one or more of the RPMs I nuked were necessary for the network configuration (qeth). The golden image boots, but cannot connect to the network. Luckily, we had already setup our controller per the book, and I can copy any RPM from the source tree (/nfs) to the golden image (/sles9golden), reboot the golden image, and re-install the rpm. I've already done this for the vlan RPM, but the golden image still will not come up on the network. Short of listing every RPM I purged, can anyone out there tell me what RPMs are necessary for qeth to run? Or is there further configuration I need to do on my golden image? (Oh boy, a chance to work on my ed(1) skills! :-) ) -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
Michael MacIsaac wrote: Matt, Or is there further configuration I need to do on my golden image? Consider reinstalling the golden image. What will it take - an hour? You won't have to format 100 - just select "new installation". This way you can be sure you will have all the necessary RPMs. Just don't touch the controller on minidisks 103+ "Mike MacIsaac" <[EMAIL PROTECTED]> (845) 433-7061 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 Mike, I'm trying to avoid doing this for one reason: the guy who has the laptop from which we installed the golden image is out of the office till Tuesday, and has taken his laptop with him. If it is possible to reinstall the golden image from the controller, I'd do that, but I'm having difficulties understanding how that might be possible. Would I be able to do something via "chroot /sles9golden" from the controller and running YaST? -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: RPMs to keep/purge after installation
Michael MacIsaac wrote: Matt, We've just done an install of a golden image of SLES9 almost completely by the "From LPAR to Virtual Servers in Two Days" book. Cool, good to hear. ... If we start getting rid of these RPMs in an attempt to make the golden image as lean as possible, will this break stuff we may need later? The neat thing about cloning is that you can afford to experiment. So try --erasing RPMs on the golden image you think you don't need, clone an image and test for a while (very often just trying to --erase some RPMs show interdependencies). If you find some RPMs are missing you can upgrade the golden image, reclone and test again. Repeat until you find the right set of RPMs. Well, I did spend some time getting rid of a bunch of RPMs on the golden image yesterday. Unfortunately for me, one or more of the RPMs I nuked were necessary for the network configuration (qeth). The golden image boots, but cannot connect to the network. Luckily, we had already setup our controller per the book, and I can copy any RPM from the source tree (/nfs) to the golden image (/sles9golden), reboot the golden image, and re-install the rpm. I've already done this for the vlan RPM, but the golden image still will not come up on the network. Short of listing every RPM I purged, can anyone out there tell me what RPMs are necessary for qeth to run? Or is there further configuration I need to do on my golden image? (Oh boy, a chance to work on my ed(1) skills! :-) ) Again, thanks in advance. -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
RPMs to keep/purge after installation
Greetings all, We've just done an install of a golden image of SLES9 almost completely by the "From LPAR to Virtual Servers in Two Days" book. The only difference is, in our case, we included the C/C++ compiler and tools packages in the install. Of course, this installed a lot of RPMs, some of which don't seem to be of all that much use to us. For example, I doubt we're going to need a DHCP daemon, or fetchmail, or dosfstools, or gpm. My question for the list involves those RPMs that are installed, that I've seen installed before on other distros, but I'm not sure we really need, i.e. the XFree86-* RPMs on a system that doesn't have X installed. Or the at RPM, when we plan on using and standardizing to cron. If we start getting rid of these RPMs in an attempt to make the golden image as lean as possible, will this break stuff we may need later? Thanks in advance. -- Matt Gourley Systems Administrator Pennsylvania State University Administrative Information Services - Infrastructure/Sysarc Rm 25 Shields Bldg., University Park, PA 16802 (814) 865-8726 [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390