Re: RH 7.2 Install Issues
Chet: You'r right - RedHat built their kernel for stand-alone installation only, requiring the use of the HMC integrated console, whereas VIF is built to provide Linux images with a 3215 console. You need a kernel with 3215 support to install under VIF. Romney On Mon, 22 Apr 2002 07:13:36 -0700 Chet Norris said: clip . Any time you see a posting from Alan Altmark, or Romney White about some aspect of VM or VIF, pay particular attention, as they're intimately familiar with the product from the inside out. Sorry Romney, I sent the reply to a prior note before reading yours. I've been talking to IBM, Endicott, and was echoing their conversation. Rob van der Heij wrote: If you don't want to recompile the driver yourself first: echo '$TCPIP' /proc/net/iucv/iucv0/username I'm not sure if I can do what you said. Not only can I not telnet into the image, but I also cannot communicate with the image through any console facility. The way I understand it RedHat assumes that the install is going to be performed from a HMC console, and VIF defines the console as a 3215 device for the image. There is some mis-match in device types that prevents me from entering any commands when logged on through VIF. = Chet Norris Marriott International,Inc. __ Do You Yahoo!? Yahoo! Games - play chess, backgammon, pool and more http://games.yahoo.com/
gcc compiler on Linux/390
Hi, On Win32 and on other Unixes some compilers have possibilities to optimize instruction scheduling in the compiler for Pentium pipeline architecture etc... Does anything like this exist in GCC, to optimize the instruction stream for G4, G5 or G6 etc processors ? Reinald
Re: MTBF
David Boyes [EMAIL PROTECTED] wrote: On Mon, Apr 22, 2002 at 04:26:39PM -0500, Holly, Jason wrote: has anyone established mean-time-between-failure numbers for linux instances running under vm? anything general would be good information. i'm curious about disk, memory or other system failures that compromise the vm instances. The record for us is about 9 months for a single Linux image. Average is about 3-4 months between reboots, depending on what's running in them -- things that suck up lots of memory like Websphere tend to shorten the lifespan of the machine by fragmenting storage. Machines that get a lot of interactive use tend to collect a few zombies after a while, so reboots become a reasonably good idea after a while. I have to say that I'm a little surprised at that recommendation. Let me see if I understand... you're saying that you need to reboot your Linux images about every 9 months because the Linux virtual memory manager has issues? I hadn't heard that before... Seems like I've heard lots of tales of people with Linux up much longer than 9 months... doing web services, etc... do you think your 9 month figure is a function of the 390 version of Linux, or Linux in general? That is, would you recommend rebooting a PC version of Linux on the same interval (given the same workload?) We don't reboot our machines unless there is some configuration change... but, we don't do highly interactive things on them, just product builds every now-and-then. Also - just to ask - what about the BSD variants - would you also recommend 9 months for them? It's my understanding that the virtual memory manager in FreeBSD is better than Linux, for example, and perhaps might not suffer the fragmentation issues? [Of course, that may be a biased understanding... I have no evidence to back that up.] For the VM side, I know of sites with uptimes measured in years; Yes - this is much better! :-) And, I would expect it from Linux (even today's Linux) too... Could you relate more about this 9 month figure? Do you have specific instances where it was required, that you can share of course... - Thanks! - - Dave Rivers - -- [EMAIL PROTECTED]Work: (919) 676-0847 Get your mainframe programming tools at http://www.dignus.com
Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00
Thanks for your response, Rob, but the website you have provided assumes OS/390 is running on the machine on which Linux is being installed. I am attempting to to run this under VM, and don't have OS/390 installed on this box. I hope all will forgive me for the following, it is the only way to convey the whole picture. (lin is a REXX that deletes any reader files, then reloads the LINUX files and executes.) parmfile: ramdisk_size=32768 dasd=200-202 root=/dev/ram0 ro ipldelay=30s CTCA 0E68 COUPLED TO TCPIP 0E69 CTCA 0E69 COUPLED TO TCPIP 0E68 Ready; T=0.48/0.56 09:11:54 lin 003 FILES PURGED RDR FILE 0103 SENT FROM LINUX1 PUN WAS 0103 RECS 019K CPY 001 A NOHOLD NOKEEP RDR FILE 0104 SENT FROM LINUX1 PUN WAS 0104 RECS 0001 CPY 001 A NOHOLD NOKEEP RDR FILE 0105 SENT FROM LINUX1 PUN WAS 0105 RECS 123K CPY 001 A NOHOLD NOKEEP 003 FILES CHANGED Linux version 2.2.16 ([EMAIL PROTECTED]) (gcc version 2.95.2 19991024 (releas e)) #1 SMP Fri Nov 3 09:38:59 GMT 2000 Command line is: We are running under VM This machine has no IEEE fpu Detected device 0E68 on subchannel - PIM = 80, PAM = 80, POM = FF Detected device 0E69 on subchannel 0001 - PIM = 80, PAM = 80, POM = FF Detected device 0191 on subchannel 0002 - PIM = 80, PAM = 80, POM = FF Detected device 0200 on subchannel 0003 - PIM = 80, PAM = 80, POM = FF Detected device 0201 on subchannel 0004 - PIM = 80, PAM = 80, POM = FF Detected device 0202 on subchannel 0005 - PIM = 80, PAM = 80, POM = FF (etc) Dected device 0120 on subchannel 000E - PIM = 80, PAM = 80, POM = FF Highest subchannel number detected (hex) : 000E SenseID : device 0E68 reports: Dev Type/Mod = 3088/08 SenseID : device 0E69 reports: Dev Type/Mod = 3088/08 SenseID : device 0191 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0200 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0201 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0202 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0009 reports: Dev Type/Mod = 3215/00 (etc.) SenseID : device 0120 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 Calibrating delay loop... 15.10 BogoMIPS SenseID : device 019D reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 SenseID : device 0198 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 SenseID : device 0120 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 Calibrating delay loop... 15.26 BogoMIPS Memory: 257080k/262144k available (1088k kernel code, 0k reserved, 3976k data, 0 k init) Dentry hash table entries: 32768 (order 6, 256k) Buffer cache hash table entries: 262144 (order 8, 1024k) Page cache hash table entries: 65536 (order 6, 256k) debug: 16 areas reserved for debugging information debug: reserved 4 areas of 4 pages for debugging ccwcache VFS: Diskquotas version dquot_6.4.0 initialized POSIX conformance testing by UNIFIX Detected 1 CPU's Boot cpu address 0 cpu 0 phys_idx=0 vers=FF ident=0D0607 machine=7490 unused= Linux NET4.0 for Linux 2.2 Based upon Swansea University Computer Society NET3.039 NET4: Unix domain sockets 1.0 for Linux NET4.0. NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP, IGMP TCP: Hash tables configured (ehash 262144 bhash 65536) Linux IP multicast router 0.06 plus PIM-SM Starting kswapd v 1.5 pty: 256 Unix98 ptys configured RAM disk driver initialized: 16 RAM disks of 32768K size loop: registered device at major 7 LVM version 0.8i by Heinz Mauelshagen (02/10/1999) lvm -- Driver successfully initialized md driver 0.36.6 MAX_MD_DEV=4, MAX_REAL=8 Partition check: Kernel panic: VFS: Unable to mount root fs on 01:00 HCPGIR450W CP entered; disabled wait PSW 000A 8005684C -Original Message- From: Rob van der Heij [mailto:[EMAIL PROTECTED]] Sent: Monday, April 22, 2002 12:06 PM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 I followed SuSE's instructions on creating an IPL tape and used that for the IPL of the LPAR. This worked fine - I did not get the Kernel panic message this time. The only reason why I chose the DASD IPL volume (using instructions that I got from another Manual) was because it was easier to IPL from DASD than from tape. Should work though. Many reported that it works for them. Maybe check that you had the latest code from http://www.rvdheij.com/linuxipl.html If you get as far as this message it looks like Linux did pick up your parm line but failed to load the initrd. Any messages in that area? Do you get the message about the ramdisk found? Rob
RH7.2 (vs other distros) Install Issues
Is there a comparison chart on the various Linux/390 distributions anywhere, showing such feature comparisons? Possibly also listing what platforms they are known to have been sucessfully installed upon (e.g. under VM/VIF vs bare iron, and on P/390, 390 software emulation, classes of real 390 hardware, etc)? Maybe compiler issues? (If memory serves, at least one has a version that will run on older hardware missing some of the newer instructions, if the correct C compiler is used. Open your home, open your heart, become a foster parent! Romney White [EMAIL PROTECTED]@VM.MARIST.EDU on 04/23/2002 05:05:09 Please respond to Linux on 390 Port [EMAIL PROTECTED] Sent by:Linux on 390 Port [EMAIL PROTECTED] To:[EMAIL PROTECTED] cc: Subject:Re: [LINUX-390] RH 7.2 Install Issues Chet: You'r right - RedHat built their kernel for stand-alone installation only, requiring the use of the HMC integrated console, whereas VIF is built to provide Linux images with a 3215 console. You need a kernel with 3215 support to install under VIF. Romney
Re: Distribution pricing comparisons
I called all three players the same day, talked to people at each that day, and had pricing from all three within a week. I'm not sure what questions I asked that were different than yours, or who I got that you didn't, but my experience with all three has been very good. I even went back around with the educational discount? question and got timely responses from all three. Robert P. Nixinternet: [EMAIL PROTECTED] Mayo Clinic phone: 507-284-0844 200 1st St. SW page: 507-255-3450 Rochester, MN 55905 In theory, theory and practice are the same, but in practice, theory and practice are different. -Original Message- From: James Melin [SMTP:[EMAIL PROTECTED]] Sent: Monday, April 22, 2002 8:27 AM To: [EMAIL PROTECTED] Subject: Re: Distribution pricing comparisons I have to take exeption to that statement, as it has been my experience that they send you a we'll get right back to you e-mail and then do not follow through. I've asked for pricing and support offerings three times now and I've been rebuffed 3 times. I'm having my IBM people pursue it for me this time.
LINUX on S390 - TAPE IPL Problem
I am trying to IPL LINUX from tape on a S390/Multiprise 3000: I am receiving the following error -, RAMDISK: Couldn't find valid RAM DISK image starting at 0. These are the three files I copied to tape for the IPL. LINUX390.IMAGE.TXT LINUX390.INITRD.TXT LINUX390.PARM.LINE Can you help?
Re: LINUX on S390 - TAPE IPL Problem
If you copied the three files in that order, then that is your problem. The IMAGE file has to be first, followed by the PARM, and the INITRD last. -Original Message- From: Rivers, James E. [SMTP:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 9:35 AM To: [EMAIL PROTECTED] Subject: LINUX on S390 - TAPE IPL Problem I am trying to IPL LINUX from tape on a S390/Multiprise 3000: I am receiving the following error -, RAMDISK: Couldn't find valid RAM DISK image starting at 0. These are the three files I copied to tape for the IPL. LINUX390.IMAGE.TXT LINUX390.INITRD.TXT LINUX390.PARM.LINE Can you help?
Anyone running Linux on zSeries in a DMZ?
If anyone is running their Linux for zSeries or Linux for S/390 in a DMZ, whether app server or firewall, please send me a brief note describing your use. The information you provide will not be made public without your consent. Thanks. Alan Altmark Sr. Software Engineer IBM z/VM Development
Re: Sendmail Perfomance
Moloko, You still have not answered the question of just how busy your S/390 CPU is when the load average is at 10. (Not 10% as you state.) A load average does not tell you anything about how busy the CPU is, only how many processes are ready to run. Try bumping your QueueLA and RefuseLA on Linux/390 to something very large, say 99, and see what happens. But really, look at and report %CPU busy, not load average. That really doesn't tell us anything. Mark Post -Original Message- From: Moloko Monyepao [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 9:35 AM To: [EMAIL PROTECTED] Subject: Sendmail Perfomance I have sendmail 8.11.6-3 installed on Redhat 7.2 (server installation) on S390(Lpar) with an IFL dedicated to this Lpar. We made the O QueueLA=10 and the O RefuseLA=12 which is the same setup as on my Intel machine. When using top to check the load the following is what I get. The O MaxDaemonChildren=100 on S390 and 300 on Intel. Intel Machine = The Load average does not go over 10% when setup for both incoming and outgoing mail. S390 Machine = The Load average goes to more than 10% and it start rejecting connections if we set it up for both incoming and outgoing mail and it gets up to about 6 to 8% if we only set it up for outgoing mail which is not a lot. Please assist Moloko
Re: Distribution pricing comparisons
That's the difference. You called them. I e-mailed them. Since this linux thing here is just an experiment at this point, I have to avoid long distance charges and a rabid accounting department. The nature of the beast. |-+ | | Nix, Robert P. | | | Nix.Robert@mayo.| | | edu | | | Sent by: Linux on| | | 390 Port | | | [EMAIL PROTECTED]| | | IST.EDU | | || | || | | 04/23/2002 07:38 | | | AM | | | Please respond to| | | Linux on 390 Port| | || |-+ --| | | | To: [EMAIL PROTECTED] | | cc: | | Subject: Re: Distribution pricing comparisons | --| I called all three players the same day, talked to people at each that day, and had pricing from all three within a week. I'm not sure what questions I asked that were different than yours, or who I got that you didn't, but my experience with all three has been very good. I even went back around with the educational discount? question and got timely responses from all three.
How do we get iostat working in SLES7?
We are having some problems at the moment geting iostat to output data from the partitons defined to the system (e.g iostat -x /dev/dasdb). This works OK on Redhat 7.2 and the reason for this seems to be that /proc/partitions contains rather more information on RedHat than SLES7. We tried applying linux-2.4.0-sard.patch to SLES7: --- linux/drivers/block/ll_rw_blk.c.~1~ Mon Jul 17 14:53:34 2000 --- linux/drivers/scsi/scsi_lib.c.~1~ Mon Jul 17 14:53:30 2000 --- linux/fs/partitions/check.c.~1~ Mon Jul 17 14:53:31 2000 --- linux/include/linux/blkdev.h.~1~Mon Jul 17 14:53:38 2000 --- linux/include/linux/genhd.h.~1~ Mon Jul 17 14:53:28 2000 --- linux/drivers/block/genhd.c.origFri Sep 7 12:05:30 2001 linux/drivers/block/ll_rw_blk.c.~1~ patches cleanly but the others do not. Are we going about this the wrong way, is there an alternative utility that provides the same level of detail. Any suggestions welome! John P Taylor ([EMAIL PROTECTED])
Re: RH7.2 (vs other distros) Install Issues
During the Distributions Redbook residency, the team installed SuSE, Red Hat, Turbolinux, and Millenux in an LPAR, under VIF, and under VM. They all worked (at that time). Since then, there have been numerous changes, and I am not aware of anyone that has gone through the same exercise since. Chet's particular problem is due to a bug that was introduced to the IUCV driver after the Redbook testing was done. No chart is ever going to be able to keep that kind of information current. Mark Post -Original Message- From: Carey Schug [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 8:19 AM To: [EMAIL PROTECTED] Subject: RH7.2 (vs other distros) Install Issues Is there a comparison chart on the various Linux/390 distributions anywhere, showing such feature comparisons? Possibly also listing what platforms they are known to have been sucessfully installed upon (e.g. under VM/VIF vs bare iron, and on P/390, 390 software emulation, classes of real 390 hardware, etc)? Maybe compiler issues? (If memory serves, at least one has a version that will run on older hardware missing some of the newer instructions, if the correct C compiler is used. Open your home, open your heart, become a foster parent! Romney White [EMAIL PROTECTED]@VM.MARIST.EDU on 04/23/2002 05:05:09 Please respond to Linux on 390 Port [EMAIL PROTECTED] Sent by:Linux on 390 Port [EMAIL PROTECTED] To:[EMAIL PROTECTED] cc: Subject:Re: [LINUX-390] RH 7.2 Install Issues Chet: You'r right - RedHat built their kernel for stand-alone installation only, requiring the use of the HMC integrated console, whereas VIF is built to provide Linux images with a 3215 console. You need a kernel with 3215 support to install under VIF. Romney
Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00
Jill, I don't know why, but the kernel is not seeing your parmline: Command line is: We are running under VM This is why you're getting the kernel panic. Could you post the contents of your lin exec, and where you got your kernel from? Mark Post -Original Message- From: Jill Grine [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 8:29 AM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 Thanks for your response, Rob, but the website you have provided assumes OS/390 is running on the machine on which Linux is being installed. I am attempting to to run this under VM, and don't have OS/390 installed on this box. I hope all will forgive me for the following, it is the only way to convey the whole picture. (lin is a REXX that deletes any reader files, then reloads the LINUX files and executes.) parmfile: ramdisk_size=32768 dasd=200-202 root=/dev/ram0 ro ipldelay=30s CTCA 0E68 COUPLED TO TCPIP 0E69 CTCA 0E69 COUPLED TO TCPIP 0E68 Ready; T=0.48/0.56 09:11:54 lin 003 FILES PURGED RDR FILE 0103 SENT FROM LINUX1 PUN WAS 0103 RECS 019K CPY 001 A NOHOLD NOKEEP RDR FILE 0104 SENT FROM LINUX1 PUN WAS 0104 RECS 0001 CPY 001 A NOHOLD NOKEEP RDR FILE 0105 SENT FROM LINUX1 PUN WAS 0105 RECS 123K CPY 001 A NOHOLD NOKEEP 003 FILES CHANGED Linux version 2.2.16 ([EMAIL PROTECTED]) (gcc version 2.95.2 19991024 (releas e)) #1 SMP Fri Nov 3 09:38:59 GMT 2000 Command line is: We are running under VM This machine has no IEEE fpu Detected device 0E68 on subchannel - PIM = 80, PAM = 80, POM = FF Detected device 0E69 on subchannel 0001 - PIM = 80, PAM = 80, POM = FF Detected device 0191 on subchannel 0002 - PIM = 80, PAM = 80, POM = FF Detected device 0200 on subchannel 0003 - PIM = 80, PAM = 80, POM = FF Detected device 0201 on subchannel 0004 - PIM = 80, PAM = 80, POM = FF Detected device 0202 on subchannel 0005 - PIM = 80, PAM = 80, POM = FF (etc) Dected device 0120 on subchannel 000E - PIM = 80, PAM = 80, POM = FF Highest subchannel number detected (hex) : 000E SenseID : device 0E68 reports: Dev Type/Mod = 3088/08 SenseID : device 0E69 reports: Dev Type/Mod = 3088/08 SenseID : device 0191 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0200 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0201 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0202 reports: CU Type/Mod = 3990/C2, Dev Type/Mod = 3390/0A SenseID : device 0009 reports: Dev Type/Mod = 3215/00 (etc.) SenseID : device 0120 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 Calibrating delay loop... 15.10 BogoMIPS SenseID : device 019D reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 SenseID : device 0198 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 SenseID : device 0120 reports: CU Type/Mod = 3880/01, Dev Type/Mod = 3370/00 Calibrating delay loop... 15.26 BogoMIPS Memory: 257080k/262144k available (1088k kernel code, 0k reserved, 3976k data, 0 k init) Dentry hash table entries: 32768 (order 6, 256k) Buffer cache hash table entries: 262144 (order 8, 1024k) Page cache hash table entries: 65536 (order 6, 256k) debug: 16 areas reserved for debugging information debug: reserved 4 areas of 4 pages for debugging ccwcache VFS: Diskquotas version dquot_6.4.0 initialized POSIX conformance testing by UNIFIX Detected 1 CPU's Boot cpu address 0 cpu 0 phys_idx=0 vers=FF ident=0D0607 machine=7490 unused= Linux NET4.0 for Linux 2.2 Based upon Swansea University Computer Society NET3.039 NET4: Unix domain sockets 1.0 for Linux NET4.0. NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP, IGMP TCP: Hash tables configured (ehash 262144 bhash 65536) Linux IP multicast router 0.06 plus PIM-SM Starting kswapd v 1.5 pty: 256 Unix98 ptys configured RAM disk driver initialized: 16 RAM disks of 32768K size loop: registered device at major 7 LVM version 0.8i by Heinz Mauelshagen (02/10/1999) lvm -- Driver successfully initialized md driver 0.36.6 MAX_MD_DEV=4, MAX_REAL=8 Partition check: Kernel panic: VFS: Unable to mount root fs on 01:00 HCPGIR450W CP entered; disabled wait PSW 000A 8005684C -Original Message- From: Rob van der Heij [mailto:[EMAIL PROTECTED]] Sent: Monday, April 22, 2002 12:06 PM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 I followed SuSE's instructions on creating an IPL tape and used that for the IPL of the LPAR. This worked fine - I did not get the Kernel panic message this time. The only reason why I chose the DASD IPL volume (using instructions that I got from another Manual) was because it was easier to IPL from DASD than from tape. Should work though. Many reported that it works for them. Maybe check that you had the latest code from http://www.rvdheij.com/linuxipl.html If you get as far as this message it looks like Linux did pick up your parm line
Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00
Hi Mark, I think I mentioned in my first email that the command missing line was the problem, but don't remember. Anyway here's the exec (very simple): /* REXX */ 'CLOSE RDR' 'PURGE RDR ALL' 'SPOOL PUNCH * RDR' 'PUNCH SUSE IMAGE A (NOH)' 'PUNCH SUSE PARM A (NOH)' 'PUNCH SUSE INITRD A (NOH)' 'CHANGE RDR ALL KEEP NOHOLD' 'IPL 00C CLEAR' Basically empties the reader, loads 3 files to the reader from the LINUX1 minidisk, then IPLs from the reader. I downloaded the install from ftp.suse.com/pub/suse/s390/suse-us-390/images. Bad choice? -Original Message- From: Post, Mark K [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 11:20 AM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 Jill, I don't know why, but the kernel is not seeing your parmline: Command line is: We are running under VM This is why you're getting the kernel panic. Could you post the contents of your lin exec, and where you got your kernel from? Mark Post
Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00
Hey Mark, Thanks so much for the new URL. I will certainly give it a shot and let you know. This really isn't making sense, and unfortunately I don't get to spend solid chunks of time to concentrate on it. It probably won't be until tomorrow... Gotta work on something that just came up. Don't we love the life of a sys prog? (Actually I really do...) Hope your day is as pretty as Georgia is today! (That sounded really odd...in too much of a hurry to be literate, but you get the gist.) -Original Message- From: Post, Mark K [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 11:57 AM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 Jill, No, not a bad choice, but there is an updated version available at ftp://ftp.suse.com/pub/suse/s390/kernel/2.2.16-iucv2/images/ I don't know that using the files there will fix your problem, but it won't hurt. Also, you commented that Rob's recipe for writing the IPL files to disk was based on the premise of having OS/390 available. I'm pretty sure that VM has the equivalent of ICKDSF available, so the method should be adaptable to that environment. (Of course, I don't think Rob or I ever thought it would be necessary under VM.) Mark Post -Original Message- From: Jill Grine [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 11:41 AM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 Hi Mark, I think I mentioned in my first email that the command missing line was the problem, but don't remember. Anyway here's the exec (very simple): /* REXX */ 'CLOSE RDR' 'PURGE RDR ALL' 'SPOOL PUNCH * RDR' 'PUNCH SUSE IMAGE A (NOH)' 'PUNCH SUSE PARM A (NOH)' 'PUNCH SUSE INITRD A (NOH)' 'CHANGE RDR ALL KEEP NOHOLD' 'IPL 00C CLEAR' Basically empties the reader, loads 3 files to the reader from the LINUX1 minidisk, then IPLs from the reader. I downloaded the install from ftp.suse.com/pub/suse/s390/suse-us-390/images. Bad choice? -Original Message- From: Post, Mark K [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 11:20 AM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 Jill, I don't know why, but the kernel is not seeing your parmline: Command line is: We are running under VM This is why you're getting the kernel panic. Could you post the contents of your lin exec, and where you got your kernel from? Mark Post
Enhanced CLAW support from UTS Global
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*= April 23, 2002 UTS Global, LLC is pleased to announce commercial grade support for the Cisco 7500 Series Router on Linux/390 (Linux on the mainframe). The commercial grade CLAW driver supports both CLAW packing protocol and CLAW protocol. The CLAW (Common Link Access to Workstation) channel protocol is used in TCP/IP environments to transport data between the mainframe and the Cisco Channel Interface Processor (CIP). CLAW packing is a Cisco proprietary enhancement to the CLAW protocol which enables the transport of multiple IP packets in a single channel operation. Substantial TCP/IP throughput improvement has been demonstrated. The UTS Global Enhanced CLAW driver transparently supports both PACKED and TCPIP (normal CLAW) interfaces. The driver first attempts a PACKED connection with the Cisco router, but will automatically attempt connection using the normal (TCPIP) CLAW protocol if the PACKED connection fails. Minor router configuration changes are required to enable the CLAW packing feature once the appropriate changes to the mainframe Linux kernel (CLAW driver) and host configurations are implemented. UTS Global has previously released a Cisco CLAW driver for Linux/390 under the UTS Global Public License. The latter product will continue to be available for non-commercial usage, pursuant to the terms and conditions of the UTS Global Public License. For more information regarding pricing, support and availability of the commercial CLAW packing driver, please contact: Douglas DeMers (408)496-4230 or send email to ([EMAIL PROTECTED]) Web address: www.utsglobal.com UTS Global is a trademark of UTS Global, LLC. Linux is a registered trademark of Linus Torvalds. All other registered trademarks belong to their respective holders. *=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*= BTW, we are looking for one or two sites with high network traffic (especially small packets) through CIP-connected Cisco 7XXX routers to thoroughly test the throughput improvements of this new driver. Any takers? Please email me directly at [EMAIL PROTECTED] Thank you! __ Douglas DeMers (408-496-4230) UTS Global, LLC [EMAIL PROTECTED]
backup software
what does anyone use for backup software for linux/390... do you use TSM under VM or a Microsoft product to backup Linux thanks Ralph Noll Systems Programmer City of Little Rock Phone (501) 371-4884 Fax (501) 371-4616 Cell (501) 590-8626 mailto:[EMAIL PROTECTED] \\\|/// \\\ ~ ~ /// ( @ @ ) ===oOOo=(_)=oOOo===
Re: backup software
their product requires os/390,mvs or z/os.e -Original Message- From: Froberg, David C [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 12:52 PM To: [EMAIL PROTECTED] Subject: Re: backup software Ralph, You should check in to Innovation's Upstream product. Upstream is great, works nicely and easily with existing mainframe infrastructure, and I believe it has a Linux agent. David Froberg No warrantee is implied or explicitly stated. Do not bend, fold, spindle, or mutilate. -Original Message- From: Noll, Ralph [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 1:46 PM To: [EMAIL PROTECTED] Subject:backup software what does anyone use for backup software for linux/390... do you use TSM under VM or a Microsoft product to backup Linux thanks Ralph Noll Systems Programmer City of Little Rock Phone (501) 371-4884 Fax (501) 371-4616 Cell (501) 590-8626 mailto:[EMAIL PROTECTED] \\\|/// \\\ ~ ~ /// ( @ @ ) ===oOOo=(_)=oOOo===
Re: backup software
True. I suspect a number of Linux/390 shops have MVS, OS/390, or z/OS with variety of dasd or tape pools that can be expoited for backup purposes. -Original Message- From: Noll, Ralph [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 2:13 PM To: [EMAIL PROTECTED] Subject:Re: backup software their product requires os/390,mvs or z/os.e -Original Message- From: Froberg, David C [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 12:52 PM To: [EMAIL PROTECTED] Subject: Re: backup software Ralph, You should check in to Innovation's Upstream product. Upstream is great, works nicely and easily with existing mainframe infrastructure, and I believe it has a Linux agent. David Froberg No warrantee is implied or explicitly stated. Do not bend, fold, spindle, or mutilate. -Original Message- From: Noll, Ralph [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 1:46 PM To: [EMAIL PROTECTED] Subject:backup software what does anyone use for backup software for linux/390... do you use TSM under VM or a Microsoft product to backup Linux thanks Ralph Noll Systems Programmer City of Little Rock Phone (501) 371-4884 Fax (501) 371-4616 Cell (501) 590-8626 mailto:[EMAIL PROTECTED] \\\|/// \\\ ~ ~ /// ( @ @ ) ===oOOo=(_)=oOOo===
Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00
Agreed, the ICKDSF on VM would be painful. Mark suggested loading from a different URL. Have loaded to my PC, but not to mainframe yet. Had to drop work on Linux for afternoon to work on another project, but I'm hoping the new files might make the difference. The 003 files changed is the result a changing all files in the rdr to KEEP NOHOLD in the REXX. -Original Message- From: Rob van der Heij [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 2:34 PM To: [EMAIL PROTECTED] Subject: Re: Subject: Kernel panic: VFS: Unable to mount root fs on 01: 00 At 17:56 23-04-02, you wrote: Also, you commented that Rob's recipe for writing the IPL files to disk was based on the premise of having OS/390 available. I'm pretty sure that VM has the equivalent of ICKDSF available, so the method should be adaptable to that environment. (Of course, I don't think Rob or I ever thought it would be necessary under VM.) Oh, deity forbid... obviously I used that to develop the stuff but the process with ICKDSF on VM was painful to do. I misunderstood the environment. The logging shows no ramdisk recognized. Either the ramdisk was corrupted during download or maybe the wrong kernel is being used (though I think the 0003 FILES CHANGED is only when you have the the RDR IPL configured in I think). Rob
Re: How do we get iostat working in SLES7?
Kernel updates from SuSE have the hooks for iostat to give info similar to Solaris or BSD. This was placed into 2.4.16 and beyond as I recall. This kernel patch was questionable in the past. The patch is also located on the web, there is a maintainer in France as I recall. Google it if you do not have the updated Kernel, but I would recommend that route as there are several things that went into 2.4.18 You might send a note to bernd or Jens at SuSE to check on updates. Regards, Jon Jon R. Doyle Sendmail Inc. 6425 Christie Ave Emeryville, Ca. 94608 (o_ (o_ (o_ //\ (/)_ (\)_ V_/_ On Tue, 23 Apr 2002, John P Taylor wrote: We are having some problems at the moment geting iostat to output data from the partitons defined to the system (e.g iostat -x /dev/dasdb). This works OK on Redhat 7.2 and the reason for this seems to be that /proc/partitions contains rather more information on RedHat than SLES7. We tried applying linux-2.4.0-sard.patch to SLES7: --- linux/drivers/block/ll_rw_blk.c.~1~ Mon Jul 17 14:53:34 2000 --- linux/drivers/scsi/scsi_lib.c.~1~ Mon Jul 17 14:53:30 2000 --- linux/fs/partitions/check.c.~1~ Mon Jul 17 14:53:31 2000 --- linux/include/linux/blkdev.h.~1~Mon Jul 17 14:53:38 2000 --- linux/include/linux/genhd.h.~1~ Mon Jul 17 14:53:28 2000 --- linux/drivers/block/genhd.c.origFri Sep 7 12:05:30 2001 linux/drivers/block/ll_rw_blk.c.~1~ patches cleanly but the others do not. Are we going about this the wrong way, is there an alternative utility that provides the same level of detail. Any suggestions welome! John P Taylor ([EMAIL PROTECTED])
Re: z/VM 3 and IFL engines - Oh No!
It is also my understanding that you have to buy another VM license to run guests under your IFL engine when you get it. The IFL VM will have to talk to the NON-IFL VM via an intra-LPAR communications technique. Alan Altmark posted that z/VM 3 can't run in IFL engines. This is not good news for us. We are currently running VM on a machine that can't run z/VM 4. We are looking at getting newer machine with an IFL engine. We wanted to do a Linux proof of concept on the new machine, on the IFL to isolate the linux workload. If z/VM 3 can't run on an IFL we'll have to upgrade VM - but after we get the new machine since our current machine can't run z/VM 4. So my question is can z/VM 3 really not run in an IFL, or is this another license issue? We already are pursuing RACF licensing and I now see we'll need a special bid for RSCS NJE and/or Passthru. Can the z/VM issue also be pursued as a licensing issue or is this one real? -- Rod Clayton KA3BHY Systems Programmer Howard County Public Schools [EMAIL PROTECTED]
Re: gcc compiler on Linux/390
Hi Reinald, see http://gcc.gnu.org/onlinedocs/gcc/S-390-and-zSeries-Options.html#S%2f390%20a nd%20zSeries%20Options Mark -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED]]On Behalf Of Reinald Verheij Sent: 23 April 2002 12:47 To: [EMAIL PROTECTED] Subject: gcc compiler on Linux/390 Hi, On Win32 and on other Unixes some compilers have possibilities to optimize instruction scheduling in the compiler for Pentium pipeline architecture etc... Does anything like this exist in GCC, to optimize the instruction stream for G4, G5 or G6 etc processors ? Reinald
Re: MTBF
On Tue, 23 Apr 2002 12:28:23 -0400, David Boyes [EMAIL PROTECTED] wrote: The record for us is about 9 months for a single Linux image. Average is about 3-4 months between reboots, depending on what's running in them -- things that suck up lots of memory like Websphere tend to shorten the lifespan of the machine by fragmenting storage. Machines that get a lot of interactive use tend to collect a few zombies after a while, so reboots become a reasonably good idea after a while. I have to say that I'm a little surprised at that recommendation. No, THIS IS NOT A RECOMMENDATION. This is a descriptive observation. The failures we see appear to be memory related, and there are some cases where if you cut interactive users loose and let them do their stuff, they create random garbage, et al. This is pretty standard stuff for lots of interactive processing sites -- clear the decks periodically even if it's not sick. Seems like I've heard lots of tales of people with Linux up much longer than 9 months... doing web services, etc... do you think your 9 month figure is a function of the 390 version of Linux, or Linux in general? No, I think its a function of how we make upgrade decisions and/or ops policy. I suspect that you could go longer, but I wanted to share a data point. The very long uptime reports tend to be for fixed workload environments... like a router that runs for years. john
Re: MTBF
The record for us is about 9 months for a single Linux image. Average is about 3-4 months between reboots, depending on what's running in them -- things that suck up lots of memory like Websphere tend to shorten the lifespan of the machine by fragmenting storage. Machines that get a lot of interactive use tend to collect a few zombies after a while, so reboots become a reasonably good idea after a while. I have to say that I'm a little surprised at that recommendation. No, THIS IS NOT A RECOMMENDATION. This is a descriptive observation. Ah - my fault - I didn't mean to mis-characterize your statements. My apologies. ... Sorry if the comment was confusing. I don't intend it to be a recommendation, just an observation of our experience. Your clarification helps a great deal! Many thanks! - Dave Rivers - -- [EMAIL PROTECTED]Work: (919) 676-0847 Get your mainframe programming tools at http://www.dignus.com
Re: LinuxWorld Article series
On Tue, 23 Apr 2002 05:32:03 +0800, John Summerfield [EMAIL PROTECTED] wrote: ... This is nothing really new. Sharing a VM system with early releases of MVS was unpleasant. I hear that it's no problem with the two in different LPARs, and that running MVS as a guest under VM works well with a surprisingly small performance hit (in the 2-3% ballpark.) -- --henry schaffer In the times when Sharing a VM system with early releases of MVS was unpleasant, IBM hadn't invented LPARs and I think Gene had just released (o r was about to release) the S/470s. MVS+VM, I was told, made the 168 comparable in performance to a 135. One of my first projects at Amdahl was supporting a product called VM/PE, a boringly named, technically cool piece of software which shared the real (UP) system between VM and MVS. S/370 achitecture is dependent on page zero and this code swapped page zeros between MVS and VM. It worked just fine for dedicated channels, nice low 1-2% overhead. When we started sharing control units and devices, things turned ugly. I do believe we used VM/PE, before MDF became available. We used to run two, occasionally three MVS systems on a 5860. - -- Cheers John Summerfield Microsoft's most solid OS: http://www.geocities.com/rcwoolley/ Note: mail delivered to me is deemed to be intended for me, for my disposition. == If you don't like being told you're wrong, be right!
Re: LinuxWorld Article series
On Wed, 24 Apr 2002 03:46:04 +0800, John Summerfield [EMAIL PROTECTED] wrote: On Tue, 23 Apr 2002 05:32:03 +0800, John Summerfield [EMAIL PROTECTED] wrote: ... This is nothing really new. Sharing a VM system with early releases of MVS was unpleasant. I hear that it's no problem with the two in different LPARs, and that running MVS as a guest under VM works well with a surprisingly small performance hit (in the 2-3% ballpark.) -- --henry schaffer In the times when Sharing a VM system with early releases of MVS was unpleasant, IBM hadn't invented LPARs and I think Gene had just released (o r was about to release) the S/470s. MVS+VM, I was told, made the 168 comparable in performance to a 135. One of my first projects at Amdahl was supporting a product called VM/PE, a boringly named, technically cool piece of software which shared the real (UP) system between VM and MVS. S/370 achitecture is dependent on page zero and this code swapped page zeros between MVS and VM. It worked just fine for dedicated channels, nice low 1-2% overhead. When we started sharing control units and devices, things turned ugly. I do believe we used VM/PE, before MDF became available. We used to run two, occasionally three MVS systems on a 5860. MDF was largely equal to the LPAR facility... VM/PE had a very elegant development name: Janus - who was the Roman God of portals, able to look two directions at the same time. It was originally written by Dewayne Hendricks and the original was very nice indeed. [Anyone feel free to correct me]. I ran across an original listing while at Amdahl and it was so much prettier then the product version. He was no longer working at Amdahl by the time I arrived. Robert Lerche was also involved, but I don't know whether he worked jointly with DH or not. john
netsaint per redbook ?
I am using the Linux on IBM zSeries and S/390 ISP/ASP Solutions to guide me and I have done everything in the chapter on installing netsaint and configuring apache and when I connect to http://localhost/netsaint I am told that: The requested URL /netsaint was not found on this server. This is my first foray into apache. The two statements in the httpd.conf per the pub are: Alias /netsaint/ /usr/local/netsaint/share/ ScriptAlias /cgi-bin/netsaint/ /usr/local/netsaint/bin/ Thanks Lionel B. Dyck, Systems Software Lead Kaiser Permanente Information Technology 25 N. Via Monte Ave Walnut Creek, Ca 94598 Phone: (925) 926-5332 (tie line 8/473-5332) E-Mail:[EMAIL PROTECTED] Sametime: (use Lotus Notes address) AIM:lbdyck
FOREIGN BANKS SWITCHING TO LINUX
MORE FOREIGN BANKS SWITCHING TO LINUX"A New Zealand bank has become the latest institution to adoptthe open-source Linux operating system. According to reports,the bank is to move all its branches to the Linux platform..."COMPLETE STORY:http://zdnet.com.com/2100-1104-887961.html Wilson Correia Gil (0xx11) 3179-7462 (0xx11) 3179-7044 - Fax Duratex S.A.
Re: netsaint per redbook ?
Lionel, Did you restart Apache after updating httpd.conf? Are the netsaint directories on the same file system as Apache's files? Mark Post -Original Message- From: Lionel Dyck [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 4:09 PM To: [EMAIL PROTECTED] Subject: netsaint per redbook ? I am using the Linux on IBM zSeries and S/390 ISP/ASP Solutions to guide me and I have done everything in the chapter on installing netsaint and configuring apache and when I connect to http://localhost/netsaint I am told that: The requested URL /netsaint was not found on this server. This is my first foray into apache. The two statements in the httpd.conf per the pub are: Alias /netsaint/ /usr/local/netsaint/share/ ScriptAlias /cgi-bin/netsaint/ /usr/local/netsaint/bin/ Thanks Lionel B. Dyck, Systems Software Lead Kaiser Permanente Information Technology 25 N. Via Monte Ave Walnut Creek, Ca 94598 Phone: (925) 926-5332 (tie line 8/473-5332) E-Mail:[EMAIL PROTECTED] Sametime: (use Lotus Notes address) AIM:lbdyck
Re: Sendmail Perfomance
[EMAIL PROTECTED] said: We have also seen large performance gains on 2.4.18, especially with the SuSE patch from Andrea (VM33). I don't know why, but disk performance improved on several Intel/AMD boxes I use by around 30% at 2.4.17. As measured by hdparm, but bonnie/bonnie++ backed that up on my Athlon system. -- Cheers John Summerfield Microsoft's most solid OS: http://www.geocities.com/rcwoolley/ Note: mail delivered to me is deemed to be intended for me, for my disposition. == If you don't like being told you're wrong, be right!
Re: Sendmail Perfomance
I think that is when Andrea's work went in to mainline. On SuSE builds it is there for all platforms. I think some of the -aa tree went into the mainline, with other pieces from Andrew. I would have to look back over the 10k lkml entries :~) You will also see some gains over the stock using .19pre7 that has a lot of work on ReiserFS done (some specific to Sendmail) as well as more cleanup from Andrea. You can grab the source at ftp.suse.de/pub.people/mantel Regards, Jon Jon R. Doyle Sendmail Inc. 6425 Christie Ave Emeryville, Ca. 94608 (o_ (o_ (o_ //\ (/)_ (\)_ V_/_ On Wed, 24 Apr 2002, John Summerfield wrote: [EMAIL PROTECTED] said: We have also seen large performance gains on 2.4.18, especially with the SuSE patch from Andrea (VM33). I don't know why, but disk performance improved on several Intel/AMD boxes I use by around 30% at 2.4.17. As measured by hdparm, but bonnie/bonnie++ backed that up on my Athlon system. -- Cheers John Summerfield Microsoft's most solid OS: http://www.geocities.com/rcwoolley/ Note: mail delivered to me is deemed to be intended for me, for my disposition. == If you don't like being told you're wrong, be right!
Samba article
http://www.itweek.co.uk/News/1131114 Samba runs rings around Windows Lionel B. Dyck, Systems Software Lead Kaiser Permanente Information Technology 25 N. Via Monte Ave Walnut Creek, Ca 94598 Phone: (925) 926-5332 (tie line 8/473-5332) E-Mail:[EMAIL PROTECTED] Sametime: (use Lotus Notes address) AIM:lbdyck
Re: Sendmail Perfomance
Hi Jon, Significant performance increases will be seen using the ReiserFS for Queues dirs due to the small random file counts. You can find all kinds of info on ReiserFS on the IBM site. The other thing you would really want to investigate is Sendmail 8.12. Yes, Yes, not just another go to the new rev 8.10 actually performs better in some cases, 8.12 has this performance plus advantages in the allocation of processes and queue definition. I do know that SuSE builds of 8.12 are available, have not checked right away on zSeries. You can always use source. We have also seen large performance gains on 2.4.18, especially with the SuSE patch from Andrea (VM33). Up to now I thought that every filesystem with journaling lowers disk performance ? The journaling function produces additional CPU load and i/o's. So I'm suprised that ReiserFS should be faster than ext2. Let's see if I find time to try it out. I just read the list of new features from sendmail 8.12. Seems that for our purpose - a smtp-relay that is mx-backup for several domains - upgrading to 8.12.3 could result in much better performance. I think it's worth building it on the zSeries. Thanks, Robert.
Re: netsaint per redbook ?
This is under SLES 7.2 and yes I did a rcapache restart. The files are in the same filesystem. I have to believe there is something in the apache setup but I don't know where to look. I also followed the info in the SHARE Linux hands-on-lab for SuSE to setup apache. And I am not running the firewall. thx Linux on 390 Port [EMAIL PROTECTED] wrote on 04/23/2002 01:55:46 PM: Lionel, Did you restart Apache after updating httpd.conf? Are the netsaint directories on the same file system as Apache's files? Mark Post -Original Message- From: Lionel Dyck [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 4:09 PM To: [EMAIL PROTECTED] Subject: netsaint per redbook ? I am using the Linux on IBM zSeries and S/390 ISP/ASP Solutions to guide me and I have done everything in the chapter on installing netsaint and configuring apache and when I connect to http://localhost/netsaint I am told that: The requested URL /netsaint was not found on this server. This is my first foray into apache. The two statements in the httpd.conf per the pub are: Alias /netsaint/ /usr/local/netsaint/share/ ScriptAlias /cgi-bin/netsaint/ /usr/local/netsaint/bin/ Thanks Lionel B. Dyck, Systems Software Lead Kaiser Permanente Information Technology 25 N. Via Monte Ave Walnut Creek, Ca 94598 Phone: (925) 926-5332 (tie line 8/473-5332) E-Mail:[EMAIL PROTECTED] Sametime: (use Lotus Notes address) AIM:lbdyck
Re: MTBF
I made the same experience with a heavy loaded nntp-server. We have to reboot the system after about 10-12 weeks. Looks like a memory leak in the kernel (2.4.7). Robert, who has a machine running solaris 2.5.1 with an uptime of 1386 (!) days ;-) - Original Message - From: David Boyes [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, April 23, 2002 6:28 PM Subject: Re: MTBF The record for us is about 9 months for a single Linux image. Average is about 3-4 months between reboots, depending on what's running in them -- things that suck up lots of memory like Websphere tend to shorten the lifespan of the machine by fragmenting storage. Machines that get a lot of interactive use tend to collect a few zombies after a while, so reboots become a reasonably good idea after a while. I have to say that I'm a little surprised at that recommendation. No, THIS IS NOT A RECOMMENDATION. This is a descriptive observation. The failures we see appear to be memory related, and there are some cases where if you cut interactive users loose and let them do their stuff, they create random garbage, et al. This is pretty standard stuff for lots of interactive processing sites -- clear the decks periodically even if it's not sick. Seems like I've heard lots of tales of people with Linux up much longer than 9 months... doing web services, etc... do you think your 9 month figure is a function of the 390 version of Linux, or Linux in general? No, I think its a function of how we make upgrade decisions and/or ops policy. I suspect that you could go longer, but I wanted to share a data point. That is, would you recommend rebooting a PC version of Linux on the same interval (given the same workload?) Given my workload, probably. My users are rude, cranky, and badly behaved. They can break anything...8-). Also - just to ask - what about the BSD variants - would you also recommend 9 months for them? See above. If the users do stupid things, you're about in the same position no matter what the OS. Could you relate more about this 9 month figure? Do you have specific instances where it was required, that you can share of course... The 9 month one was a power failure on site with a P390 doing mail delivery. The failure was that the customer was too ... funds limited... to go for backup setups. Restart, and we were up and running, but that's the maximum runtime we've observed. Sorry if the comment was confusing. I don't intend it to be a recommendation, just an observation of our experience. -- db
try to load Gigabit OSA Ethernet driver qeth
I have SuSE v7 running on an S390 LPAR. I tried to load Gigabit OSA Ethernet driver with the following command: insmod qeth qeth_options=noauto,0x0f0a,0x0f0b,0x0f0c This command failed with the following messages: Using /lib/modules/2.2.16/net/qeth.o /lib/modules/2.2.16/net/qeth.o: init_module: Device or resource busy Hint: this error can be caused by incorrect module parameters, including invalid IO or IRQ parameters Any suggestions what I did wrong? Thanks. IC Liang.
Re: Samba article
http://www.itweek.co.uk/News/1131114 Samba runs rings around Windows So says http://www.dwheeler.com/oss_fs_why.html - but it takes a while to read. -- Cheers John Summerfield Microsoft's most solid OS: http://www.geocities.com/rcwoolley/ Note: mail delivered to me is deemed to be intended for me, for my disposition. == If you don't like being told you're wrong, be right!
Re: Sendmail Perfomance
used in the Sendmail deployments we do, as the Queues or mailstores can become corrupt, especially with the large Cache on today's controllers. EXT2 does not fsync or dirsync correctly, we had to place patches into 8.12 code base for this problematic issue (people that use EXT2 anyway). This is incorrect. Ext2 is strictly standards compliant here. Sendmail at least used to make totally bogus assumptions about synchronous directory updating. I'm suprised that you might not have fixed such code. (Perhaps you have - I run exim because I want fast efficient mail handling and a human readable configuration) EXT3 will have the same set of problems, infact I was just doing studies Also incorrect. See the documentation. Ext3 also follows the posix/sus requirements and allows you to select additional ordering guarantees As to reiserfs being very fast for very small files. It was certainly designed to be , and Namesys are smart people.
Re: MTBF
I made the same experience with a heavy loaded nntp-server. We have to reboot the system after about 10-12 weeks. Looks like a memory leak in the kernel (2.4.7). 2.4.7 has dcache and other vm balancing problems. Thats one of the reasons its considered obsolete.
Re: try to load Gigabit OSA Ethernet driver qeth
On Tue, Apr 23, 2002 at 03:12:27PM -0700, Liang, Ih-Cheng wrote: I have SuSE v7 running on an S390 LPAR. I tried to load Gigabit OSA Ethernet driver with the following command: insmod qeth qeth_options=noauto,0x0f0a,0x0f0b,0x0f0c This command failed with the following messages: Using /lib/modules/2.2.16/net/qeth.o /lib/modules/2.2.16/net/qeth.o: init_module: Device or resource busy Hint: this error can be caused by incorrect module parameters, including invalid IO or IRQ parameters This sounds like the module has already been loaded. Try lsmod to see if you see it loaded already and rmmod it if it is. -- David Rock [EMAIL PROTECTED]
Re: Sendmail Perfomance
On Wed, 24 Apr 2002, Alan Cox wrote: used in the Sendmail deployments we do, as the Queues or mailstores can become corrupt, especially with the large Cache on today's controllers. EXT2 does not fsync or dirsync correctly, we had to place patches into 8.12 code base for this problematic issue (people that use EXT2 anyway). This is incorrect. Ext2 is strictly standards compliant here. Sendmail at least used to make totally bogus assumptions about synchronous directory updating. I'm suprised that you might not have fixed such code. (Perhaps you have - I run exim because I want fast efficient mail handling and a human readable configuration) Whether or not it meets some standards these issues were problematic, and only on Linux, not BSD with Soft Updates, or the other likely suspects on Unix AIX etc. Whn I uncovered this, the Opensource group did make fixes specifically for Linux, it is in the release notes for 8.12 Quite standard with some mail systems, there are threads on this issue in the SpecBench area that handles Specmail. This was really when it came to a head for us. In fact Domino was rejected until they chattr or mounted EXT2 in -sync. From my memory when I spoke to Claus and Greg here the ISV being required to make specific calls to a FS was hard to swallow. However, we did do this in 8.12 EXT3 will have the same set of problems, infact I was just doing studies Also incorrect. See the documentation. Ext3 also follows the posix/sus requirements and allows you to select additional ordering guarantees Humm, if you mean that you can mount EXT3 with options, then yes, but I rarely find folks that follow the warning label. If you mean the forced fsync, yes, that too can be disabled, but I have not read wheter this is wise just saw in places that you should still run fsync manually at times. This last area was specific to HA environments. As to reiserfs being very fast for very small files. It was certainly designed to be , and Namesys are smart people. Yes, smart, Hans just needs better PR for hot technology or maybe RH should put ReiserFS support in their installer then all the American's would use it, since we buy and use everything from Software to Wonder Bread on marketing perception :~) Regards, Jon
Re: Sendmail Perfomance
In fact Domino was rejected until they chattr or mounted EXT2 in -sync. From my memory when I spoke to Claus and Greg here the ISV being required to make specific calls to a FS was hard to swallow. However, we did do this in 8.12 Standards exist for a reason. If there really is a problem with lack of standards here it would be very good if Sendmail Inc raised it with the Open Group for SuSv4. I support the need for a standard to handle the directory name synchronizing. Humm, if you mean that you can mount EXT3 with options, then yes, but I rarely find folks that follow the warning label. If you mean the forced fsync, yes, that too can be disabled, but I have not read wheter this is wise just saw in places that you should still run fsync manually at times. This last area was specific to HA environments. No I mean that it defines orderings and those orderings can handle buggy software that assumes ancient 4BSD unix behaviour.
Re: try to load Gigabit OSA Ethernet driver qeth
Did you load the qdio module first? Mark Post -Original Message- From: Liang, Ih-Cheng [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 6:12 PM To: [EMAIL PROTECTED] Subject: try to load Gigabit OSA Ethernet driver qeth I have SuSE v7 running on an S390 LPAR. I tried to load Gigabit OSA Ethernet driver with the following command: insmod qeth qeth_options=noauto,0x0f0a,0x0f0b,0x0f0c This command failed with the following messages: Using /lib/modules/2.2.16/net/qeth.o /lib/modules/2.2.16/net/qeth.o: init_module: Device or resource busy Hint: this error can be caused by incorrect module parameters, including invalid IO or IRQ parameters Any suggestions what I did wrong? Thanks. IC Liang.
Uncensored Redbooks - Revisited
Well, so far I haven't been able to line up anyone willing to host the uncensored version of the Linux/390 Redbook. If anyone on the mailing list is able and willing to do that, please let me know. The file is about 5MB in size. If someone wants to host the book, but doesn't have a copy of the original, I can provide that. Thanks, Mark Post -Original Message- From: Post, Mark K Sent: Friday, April 19, 2002 4:49 PM To: 'Linux390' Subject: Uncensored Redbooks I've gotten permission from my web hoster to put a link to an uncensored copy of SG24-4987 on the main page of linuxvm.org. I will be updating the main page very soon to point to it, along with a short explanation of why there are two pointers. Mark Post
Re: LinuxWorld Article series
Mark, This is an Amdahl Millennium 700 which is ALS-2 compliant but does NOT have IEEE. At 09:58 AM 4/22/02 -0400, you wrote: David, No, I've been informed by a reliable source that this is an MSF'd Amdahl 0700 processor. Mark Post -Original Message- From: David Boyes [mailto:[EMAIL PROTECTED]] Sent: Monday, April 22, 2002 8:29 AM To: [EMAIL PROTECTED] Subject: Re: LinuxWorld Article series If I'm reading it correctly, a 6070 is some kind of PowerPC box. Possibly a R/390? If so, you're facing the same OS/2 based device emulation... -- db Dave, Not really, sorry. I'm just a user there, and it sits in Texas. I can ask the VM guy that supports it if you're really curious. Mark Post -Original Message- From: Dave Jones [mailto:[EMAIL PROTECTED]] Sent: Saturday, April 20, 2002 9:38 PM To: [EMAIL PROTECTED] Subject: Re: LinuxWorld Article series Mark, I don't recognize the CPU type in the CPUID field. can you explain what type of system you ran this test on? Thanks. DJ CPUID = FF0240760700 Jon Nolting (925) 672-1249 -- Home office
Re: Sendmail Perfomance
Mark! The %CPU goes up to 60 to 100 when the load gets high. The following is some of the stats I got from TOP. CPU states: 96.4% user, 3.5% system, 0.0% nice, 0.0% idle Load= 8 CPU states: 68.1% user, 2.5% system, 0.0% nice, 29.2% idle Load=10 I want to increase the load to maybe 30 or 40 but what I want to know is whether or not will I get connection rejections because looking at the stats as it is now is with only Outgoing mail I get an impression that I will have problems when I increase the load and accept both incoming and outgoing mail. Please Advice Moloko -Original Message- From: Post, Mark K [SMTP:[EMAIL PROTECTED]] Sent: 23 April 2002 16:44 To: [EMAIL PROTECTED] Subject: Re: Sendmail Perfomance Moloko, You still have not answered the question of just how busy your S/390 CPU is when the load average is at 10. (Not 10% as you state.) A load average does not tell you anything about how busy the CPU is, only how many processes are ready to run. Try bumping your QueueLA and RefuseLA on Linux/390 to something very large, say 99, and see what happens. But really, look at and report %CPU busy, not load average. That really doesn't tell us anything. Mark Post -Original Message- From: Moloko Monyepao [mailto:[EMAIL PROTECTED]] Sent: Tuesday, April 23, 2002 9:35 AM To: [EMAIL PROTECTED] Subject: Sendmail Perfomance I have sendmail 8.11.6-3 installed on Redhat 7.2 (server installation) on S390(Lpar) with an IFL dedicated to this Lpar. We made the O QueueLA=10 and the O RefuseLA=12 which is the same setup as on my Intel machine. When using top to check the load the following is what I get. The O MaxDaemonChildren=100 on S390 and 300 on Intel. Intel Machine = The Load average does not go over 10% when setup for both incoming and outgoing mail. S390 Machine = The Load average goes to more than 10% and it start rejecting connections if we set it up for both incoming and outgoing mail and it gets up to about 6 to 8% if we only set it up for outgoing mail which is not a lot. Please assist Moloko
Re: Sendmail Perfomance
Robert! According to our monitoring we did not notice any perfomance problems on the disks or the network. The disk is using Raid0 on about 4 disk which can be reached by about (4 ficon channels and 4 escon channels shared between 5 Lpars). Thanx Moloko -Original Message- From: Robert Werner [SMTP:[EMAIL PROTECTED]] Sent: 23 April 2002 16:41 To: [EMAIL PROTECTED] Subject: Re: Sendmail Perfomance Hi Moloko, We made the experience that you cannot compare the load averages directly between Intel an Linux/390. Also the shown load average on our linuxes is higher using a 2.4 kernel as with kernel 2.2.x (why?) But if there's really a problem with the performance on the S390 you have to find out where your bottleneck is. On our smtp-relay on the z900 (under VM, SuSE7.0 with kernel 2.4.7, sendmail 8.11) the disk performance slowed down the system and pushed up the load average. A solution is to attach the disks using more channels and to stripe the spool-area over several disks and channels with LVM or software Raid0. Sendmail also slows down if the number of mails in your spooldir gets to high. So if you have a huge amount of spooled mails it is better to split the spool area in several subdirs with dedicated sendmail instances processing the queues. We also had bad performance with kernel 2.2.x. Robert. - Original Message - From: Moloko Monyepao [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, April 23, 2002 3:35 PM Subject: Sendmail Perfomance I have sendmail 8.11.6-3 installed on Redhat 7.2 (server installation) on S390(Lpar) with an IFL dedicated to this Lpar. We made the O QueueLA=10 and the O RefuseLA=12 which is the same setup as on my Intel machine. When using top to check the load the following is what I get. The O MaxDaemonChildren=100 on S390 and 300 on Intel. Intel Machine = The Load average does not go over 10% when setup for both incoming and outgoing mail. S390 Machine = The Load average goes to more than 10% and it start rejecting connections if we set it up for both incoming and outgoing mail and it gets up to about 6 to 8% if we only set it up for outgoing mail which is not a lot. Please assist Moloko