Re: signal command for shutdown - Howto?
On Mon, Oct 25, 2004 at 09:12:44AM -0500, James Melin wrote: > Is there something within linux that you have to do to tell it what you > expect it to do when it gets that signal? > Yeah, that's where you add the ctrlaltdel line to /etc/inittab, IIRC. Some distributions already have this done for you. 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
zTPF
See: http://linuxtoday.com/high_performance/2004102501526NWHESW "Users of Transaction Processing Facility (TPF), IBM's mainframe operating system, seem to have both pride in the OS and an inferiority complex, often referring to it as 'IBM's unknown operating system.' Even IBM called it 'little known' in a press release for its latest version of the system, zTPF. "IBM in early October announced that it was taking a step toward bringing zTPF to the attention of the next generation of IT workers. With a planned release date of September 30, 2005, zTPF adds features such as 64-bit addressing capability on zSeries servers, but what is most significant is the system's compatibility with Linux. This feature opens the door for many more programmers and administrators who may want to work with it. -- 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: CAN-2004-0887
Martin Schwidefsky wrote: CAN-2004-0887 is a local root exploit specific to s390. The only affected distro is SLES9 and they have a security update in place. Do you need to know anything more specific ? Thanks Martin. I was curious how the exploit worked. Your patch adds a new handler for program interrupt 0x1c (space-switch) where you make sure a saved problem state psw is in home space mode: + regs->psw.mask |= PSW_ASC_HOME; So it looks like problem state (user?) code should always run in home space mode. If sacf is issued to switch out of home space mode then the psw is updated accordingly, however if the space-switch-event is on in cr1 or cr13 then a 0x1c interrupt will occur (looks like this bit is always on in cr13). Before the patch, a problem state program could issue sacf to get out of home space mode and then a 0x1c program interrupt occurs and do_trap(SIGILL) is performed but regs->psw has the wrong address-space-control bits. I presume, then, that a signal handler could be called, but instead of code getting executed in home space, code would get executed in primary space instead. If a carefully crafted signal handler address was created then the code actually executed could put the user space in root mode ?? I suppose what I am really trying to understand a little better is how s390 linux works. This is what I'm guessing: 1) userland runs in home space mode 2) kernel runs in primary space mode, uses mvcs/mvcp to copy between kernel and userland 3) syscall is implemented using the svc instruction 4) cow is implementing by forcing program interrupt 0x04 on write Greg Smith -- 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
New England Users of VM (NEUVM) Autumn Meeting - November 9, 2004
<> Hi Folks, Reminder, if you are planning on attending the NEUVM Autumn Meeting please take a moment to register NOW. :) NEUVM will be holding it's Autumn Meeting on November 9, 2004 at Lombardo's in Randolph, Massachusetts. For meeting details and online registration click here: http://www.neuvm.org/meetings.php Once again, registration is FREE and includes a breakfast buffeet and fabulous Lombardo's luncheon, featuring your choice of Roast Prime Rib of Beef or New England Roasted Turkey Dinner! Advance online registration is required, and we need to determine a head-count for meals so please register ASAP. Our meeting theme is "Disaster Recovery for z/VM and Linux for z/Series" - and we've got some great presentations scheduled. Presentations include: IBM's LINUX Strategy - Leonard Santalucia, IBM Len will be discussing IBM's Linux Strategy and Directions from a business, technical, and financial perspective using customer examples including disaster recovery scenarios. Automating z/VM and Linux for z/Series Disaster Recovery On A Budget - Michael Coffin, MC Consulting Company, Inc. Michael Coffin will discuss options and methods for automating z/VM and Linux for z/Series Disaster Recovery backup and recovery using IBM's DASD Dump Restore (DDR) program. Included in the discussion will be use of Standard Label tape processing, stacking multiple DASD volumes on high capacity tapes, interfacing with DFSMS/RMS and VM:Tape, minimizing service disruptions using Ramac SnapShot technology and how to get up and running FAST at the Disaster Recovery site with minimal human intervention using untrained staff. VDisaster Recovery, Backup & Restore for z/VM & Linux - John Hall, Safe Software The emergence of Linux/390 on z/VM has brought new interest to z/VM. One of the first tools you need when deploying Linux on the mainframe is a reliable disaster recovery and backup/restore solution. Historically, z/VM installations have used expensive vendor solutions or complicated and difficult to maintain DDR implementations. This session covers why you need to have a DR plan, how to implement DR and backup/restore with DDR, and an overview of Safe Software's newest product: SafeDR, a cost effective, turnkey z/VM disaster recovery and backup/restore solution. z/VM Platform Update - Reed Mullen, IBM z/VM Version 5.1 is now available and "open for business". This session will highlight z/VM V5.1 function announced on April 7, 2004 and added to the product since then. A more detailed perspective of future z/VM enhancements will be offered along with the latest news regarding zSeries server enhancements. Looking forward to seeing you all at Lombardo's for NEUVM's Autumn Meeting! Michael Coffin New England Users of VM (NEUVM) 289 Park Street - PMB 123 Stoughton, Massachusetts 02072 Voice: (781) 344-9837FAX: (781) 344-7683 [EMAIL PROTECTED] www.NEUVM.org -- 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: Cloneing Linux Guests on FCP SCSI
Is this a practical proceedure? What are other people doing? are they just doing new installs, and no cloning? -Cameron -Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] Behalf Of Robert J Brenneman Sent: Monday, October 25, 2004 10:20 To: [EMAIL PROTECTED] Subject: Re: Cloneing Linux Guests on FCP SCSI Well - if you're IPLing off SCSI, there is one thing you should be aware of: The Z SCSI IPL code writes information into the IPL text of the load device. This includes the FCP Device number, WWPN, and LUN. For the IPL to work, this info cannot change between when the IPL text was written and when it is actually IPLed. For example - you cannot IPL a linux from FCP device 1A00, WWPN 50050762, LUN 5001 when the IPL text was written using device 1A05. This also means that when you clone the LUN with copy services, the new LUN number will be different, and it probably ( I haven't actually tested this, but Im sure it's the case ) won't IPL. So - the Cloning will probably have to be done under Linux itself. A recipe like this should work pretty well: From a running Linux system, set up the SCSI system so that it can see the source lun A, and the target lun B use dd to copy from the source to the target: dd if=/dev/sda of=/dev/sdb mount the target somewhere chroot into the target system edit the zipl and scsi configuration files for the target system. might as well edit the network config while youre at it... make a new initrd with mkinitrd run zipl to write the new IPL text out to disk. exit chroot unmount source and target target is now IPLable You could also use copy services instead of dd for that step - it will be a little faster and use less Z cpu - but the rest of the sequence is the same. Jay Brenneman Linux Test and Integration Center T/L: 295 - 7745 Extern: 845 - 435 - 7745 [EMAIL PROTECTED] "Seader, Cameron" <[EMAIL PROTECTED] er.com>To Sent by: Linux on [EMAIL PROTECTED] 390 Port cc <[EMAIL PROTECTED] IST.EDU> Subject Cloneing Linux Guests on FCP SCSI 25/10/2004 10:51 AM Please respond to Linux on 390 Port Greetings, We are migrating all systems to FCP SCSI, and before when we were under DASD we could do cloning really easy with ddr. Is there a way now to clone when you are under FCP SCSI? is there some doc out there that has a howto for this? I guess its pretty much the same except you have to use copy services, and then somehow you have to change the FCP map within the guest. Any ideas? -Cameron Seader This transmission may contain information that is privileged, confidential and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you. A1. -- 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 [INFO] -- Access Manager: This transmission may contain information that is privileged, confidential and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you. A2 -- 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: signal command for shutdown - Howto?
On Mon, 25 Oct 2004 08:00:13 -0500, James Melin <[EMAIL PROTECTED]> wrote: > This works, but I don't know that I like having a logged on console device > just 'sitting there'. He wants to use it for some sort of automation > widget within VM to shutdown or control the guests, but I'm not really > comfortable with the exposure. Mebbey I'm paranoid but I think not in > this case. Anyone who can logon to the virtual machine can break your system. Some people rather grant root access to those with physical access than have many people know root passwords. If you can not control logon with network access, you may neen a proper External Security Manager like RACF/VM. Your VM colleage can do nice things with automated operator reading along with the console etc. But as said, if it's only for shutdown then you can do without root access. -- Rob van der Heij rvdheij @ gmail.com -- 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: Cloneing Linux Guests on FCP SCSI
Well - if you're IPLing off SCSI, there is one thing you should be aware of: The Z SCSI IPL code writes information into the IPL text of the load device. This includes the FCP Device number, WWPN, and LUN. For the IPL to work, this info cannot change between when the IPL text was written and when it is actually IPLed. For example - you cannot IPL a linux from FCP device 1A00, WWPN 50050762, LUN 5001 when the IPL text was written using device 1A05. This also means that when you clone the LUN with copy services, the new LUN number will be different, and it probably ( I haven't actually tested this, but Im sure it's the case ) won't IPL. So - the Cloning will probably have to be done under Linux itself. A recipe like this should work pretty well: >From a running Linux system, set up the SCSI system so that it can see the source lun A, and the target lun B use dd to copy from the source to the target: dd if=/dev/sda of=/dev/sdb mount the target somewhere chroot into the target system edit the zipl and scsi configuration files for the target system. might as well edit the network config while youre at it... make a new initrd with mkinitrd run zipl to write the new IPL text out to disk. exit chroot unmount source and target target is now IPLable You could also use copy services instead of dd for that step - it will be a little faster and use less Z cpu - but the rest of the sequence is the same. Jay Brenneman Linux Test and Integration Center T/L: 295 - 7745 Extern: 845 - 435 - 7745 [EMAIL PROTECTED] "Seader, Cameron" <[EMAIL PROTECTED] er.com>To Sent by: Linux on [EMAIL PROTECTED] 390 Port cc <[EMAIL PROTECTED] IST.EDU> Subject Cloneing Linux Guests on FCP SCSI 25/10/2004 10:51 AM Please respond to Linux on 390 Port Greetings, We are migrating all systems to FCP SCSI, and before when we were under DASD we could do cloning really easy with ddr. Is there a way now to clone when you are under FCP SCSI? is there some doc out there that has a howto for this? I guess its pretty much the same except you have to use copy services, and then somehow you have to change the FCP map within the guest. Any ideas? -Cameron Seader This transmission may contain information that is privileged, confidential and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you. A1. -- 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
Re: Shared filesystems with SuSE SLES 8
> I am trying to implement shared filesystems, so that multiple guests can > share one /bin and /sbin for example. ... Robert, in my experience (and from non-Linux precedent from years ago) it's much easier to share filesystems like /usr and /opt. Rob mentioned that in painfully too many cases "only part of the code is shared and some is not", so even for these two filesystems you're not always going to cover all files that might be in any given RPM. You will have to be aware. In practice, I share /usr read-only among several zSeries penguins. These instances are VERY stable in so far as they don't change much. And I've done a lot of manual twiddling to reduce what is in /bin (moving things carefully to /usr/bin). Likewise for /lib. My advice: do it! Require support for it of your Linux distributor. Require sharing tolerance from your various package vendors. You will NOT immediately recognize a lot of disk savings (and disk is cheap), but in time. You will not immediately recognize sysadmin time savings, but that too in time. But you will have to ramp up on the curve before you recognize any return on this investment. -- R; -- 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: signal command not shutting down cleanly
On Mon, 25 Oct 2004 08:03:16 -0400, Levy, Alan <[EMAIL PROTECTED]> wrote: > So what would amount of time would you recommend on the signal command ? On a thrashing system you can not predict anything, I guess. My approach would be to 'force' a few of them together and do the next when one is reported to logoff. -- Rob van der Heij rvdheij @ gmail.com -- 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: virtual hipersockets
I have just completed setting up a hipersocket connection to zVM on a Z800. We are not running Linux under VM, we have an IFL processor and Linux running in an LPAR. I have some documentation that I created for myself. I also setup an explicit route on Linux so that I can bypass the intranet and use the hipersocket for any communications to the VM system. If you would like the documentation email me at [EMAIL PROTECTED] and I will email it back to you. I tried to send it but the service said the message was to long. Good Luck Mike Lovins System Support/LAN Assiciate Consultant Ciy of Tulsa, Tulsa, Oklahoma >>> [EMAIL PROTECTED] 10/25/04 03:20AM >>> You configuration looks : world - qdio osa card - vm/TCPIP - hypersocket - LINUX. Mark , Pieter and Alan already outlined that hipersocket is not a p2p (but this is probably not the problem in this case) and pointed to verify routing problem both from linux to the world and viceversa. I only suggest to verify the routing into the osa card. OSA cards route only registered addresses to their unit addresses, and unregistered addresses to PRIMARY ROUTER (if defined) . So : - define [EMAIL PROTECTED] as PRIROUTER (if none is already defined) - OR try to use vmTCPIP as a PROXYARP (not sure you will be able, vmTCPIP ifconfig sees the link as a point2point , but the profile definition is not for a p2p, it is for an hypersocket !). hoping I didn't understand the configuration, and it helps, Franco Mignogna , IBM Italy, IT specialist -- 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
Re: new IFL, question on HMC
I have a z800 with 1 IFL processor. in the panel for (Customize Image Profiles) I show an entry in the pull down for MODE that is (Linux Only). You must select this. NOTE: refer to page 3-60 in GA22-7236-07 >>> [EMAIL PROTECTED] 10/25/04 09:37AM >>> Hi, We just had two IFL engines installed on our z900 machine. When we look at the CPs defined, I can see the standard engines, and two engines that show CF on the icons. When we go the activation profile for the LPAR that we are going to run Linux in, I do not see any LINUX options for the CPs, just the dedicate for standard engines, and the dedicate for coupling facility engines. Do I just use this CF option, or should I also have a Linux CP option? Thanks, Ken Vance Amadeus -- 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
Cloneing Linux Guests on FCP SCSI
Greetings, We are migrating all systems to FCP SCSI, and before when we were under DASD we could do cloning really easy with ddr. Is there a way now to clone when you are under FCP SCSI? is there some doc out there that has a howto for this? I guess its pretty much the same except you have to use copy services, and then somehow you have to change the FCP map within the guest. Any ideas? -Cameron Seader This transmission may contain information that is privileged, confidential and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you. A1. -- 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
Fw: Antivirus for Domino on SLES8
Aristarc, You can find a list of vendor products for Linux on zSeries at http://www-1.ibm.com/servers/eserver/zseries/solutions/s390da/linuxproduct.html Take a look at TrendMicro. They have an AV solution that plugs into Domino on Linux for zSeries. Regards, Peter Linux for zSeries Security Strategy/Design IBM System Integrity Competency Center zSeries Software System Design Group Date:Wed, 20 Oct 2004 13:41:32 +0200 From:Aristarc Diez <[EMAIL PROTECTED]> Subject: Antivirus for Domino on SLES8 We're migrating our mail servers to Domino on Linux 390 ( zSeries). Does anyone know a good antivirus for Domino? TIA ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: -- 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: new IFL, question on HMC
ken, On our z800, the IFL shows up as a CF. -- John McKown Senior Systems Programmer UICI Insurance Center Information Technology This message (including any attachments) contains confidential information intended for a specific individual and purpose, and its' content is protected by law. If you are not the intended recipient, you should delete this message and are hereby notified that any disclosure, copying, or distribution of this transmission, or taking any action based on it, is strictly prohibited. > -Original Message- > From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On > Behalf Of Ken Vance > Sent: Monday, October 25, 2004 9:38 AM > To: [EMAIL PROTECTED] > Subject: new IFL, question on HMC > > > Hi, > > We just had two IFL engines installed on our z900 machine. > When we look > at the CPs defined, I can see the standard engines, and two > engines that > show CF on the icons. When we go the activation profile for > the LPAR that > we are going to run Linux in, I do not see any LINUX options > for the CPs, > just the dedicate for standard engines, and the dedicate for coupling > facility engines. Do I just use this CF option, or should I > also have a > Linux CP option? > > Thanks, > > Ken Vance > Amadeus > > -- > 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
new IFL, question on HMC
Hi, We just had two IFL engines installed on our z900 machine. When we look at the CPs defined, I can see the standard engines, and two engines that show CF on the icons. When we go the activation profile for the LPAR that we are going to run Linux in, I do not see any LINUX options for the CPs, just the dedicate for standard engines, and the dedicate for coupling facility engines. Do I just use this CF option, or should I also have a Linux CP option? Thanks, Ken Vance Amadeus -- 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: getting a modprobe message
Well, since /usr/src/linux/Documentation/devices.txt says that char major 10 are "Non-serial mice, misc features," I suppose you can do without it. echo alias char-major-10-224 off >> /etc/modules.conf ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: -- 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: getting a modprobe message
Add this line to /etc/modules.conf alias char-major-node-10-224off Hope this helps ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: -- 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
getting a modprobe message
Oct 25 09:15:00 calhoun modprobe: modprobe: Can't locate module char-major-10-224 This is popping up every little while on the console of this machine. Should I be concerned? Not even sure why it's looking for it. -- 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: signal command for shutdown - Howto?
/etc/inittab needs an entry to handle the signal: # CTLR-ALT-DEL (signal shutdown on S/390) ca::ctrlaltdel:/sbin/shutdown -h -t 4 now -Original Message- Is there something within linux that you have to do to tell it what you expect it to do when it gets that signal? -- 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: signal command for shutdown - Howto?
Is there something within linux that you have to do to tell it what you expect it to do when it gets that signal? Adam Thornton <[EMAIL PROTECTED] mine.net> To Sent by: Linux on [EMAIL PROTECTED] 390 Port cc <[EMAIL PROTECTED] IST.EDU> Subject Re: signal command for shutdown - Howto? 10/25/2004 08:19 AM Please respond to Linux on 390 Port <[EMAIL PROTECTED] IST.EDU> On Mon, Oct 25, 2004 at 08:00:13AM -0500, James Melin wrote: > Is there a how to out there to set up VM and Linux to respond to the 'CP > SIGNAL SHUTDOWN USER' command? Our VM guy is suggesting that I make the > console TTY just default to logged on. It's a kernel option. Recent SuSE and Debian "Sarge" (don't remember about "Woody") have the option built in to respect Signal Shutdown. Don't know about Red Hat or Slackware, etc. 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 -- 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: signal command for shutdown - Howto?
On Mon, Oct 25, 2004 at 08:00:13AM -0500, James Melin wrote: > Is there a how to out there to set up VM and Linux to respond to the 'CP > SIGNAL SHUTDOWN USER' command? Our VM guy is suggesting that I make the > console TTY just default to logged on. It's a kernel option. Recent SuSE and Debian "Sarge" (don't remember about "Woody") have the option built in to respect Signal Shutdown. Don't know about Red Hat or Slackware, etc. 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
Re: z/VM LPAR Configuration Question
Hi Peter,
You should answer Linux if you plan to run under the IFL engine, VM or
not. If you answer ESA/390, your LPAR(s) will run under the "Legacy"
engine(s), VM or not again.
Notice that the options that are persented to you are not directly related
to any Operating Systems except maybe for TPF which is actually an
Operating System. For exemple, ESA/390 is valid for z/OS and OS/390, z/VM
and VM/ESA, VSE/ESA and yes Linux or any other OS that may exist and any
of those in LPAR mode. Coupling Facility is for Parallel Sysplex that is
used only with z/OS or OS/390.
Good luck.
Jean-Pierre Baril
Specialiste en technologies de l'information/IT Specialist
Novipro Inc / 2055, rue Peel, bureau 701
Montreal (Quebec) H3A 1V4
Tel: (514) 744-5353 #244, Cel: (514) 891-6848, Fax: (514) 744-3908
[EMAIL PROTECTED] / www.novipro.com
"Peter E. Abresch Jr. - at Pepco" <[EMAIL PROTECTED]>
Sent by: Linux on 390 Port <[EMAIL PROTECTED]>
2004-10-24 10:53
Please respond to
Linux on 390 Port <[EMAIL PROTECTED]>
To
[EMAIL PROTECTED]
cc
Subject
z/VM LPAR Configuration Question
We are running 2 SuSE Linux 8.1 SP03 in native LPARs using a new IFL. On
the HMC, the Linux LPAR profiles specify âLINUX onlyâ under the general
tab. We then select âNot dedicated integrated facility for Linuxâ to allow
the two Linux LPARs to share the IFL.
We are now getting ready to install z/VM V5.1. z/VM is new to us. We
defined a new LPAR just for z/VM. Our intention is to get z/VM up and
running and then migrate the two native Linux LPARs to z/VM guests.
What mode do I select for the VM LPAR profile under the general tab on the
HMC? My choices are:
ESA/390
ESA/390 TPF
Coupling Facility
Linux Only
There is no VM option. If I specify ESA/390, I do not have an option under
the processor tab to allow the VM LPAR to share the IFL. This option only
appears when I select âLinux onlyâ. We do not have excess capacity on the
central processors as the z/OS systems are fully utilizing them. Do I
define the VM LPAR as Linux only?
Thanks for any insight.
Peter
This Email message and any attachment may contain information that is
proprietary, legally privileged, confidential and/or subject to copyright
belonging to Pepco Holdings, Inc. or its affiliates ("PHI"). This Email
is
intended solely for the use of the person(s) to which it is addressed. If
you are not an intended recipient, or the employee or agent responsible
for
delivery of this Email to the intended recipient(s), you are hereby
notified that any dissemination, distribution or copying of this Email is
strictly prohibited. If you have received this message in error, please
immediately notify the sender and permanently delete this Email and any
copies. PHI policies expressly prohibit employees from making defamatory
or offensive statements and infringing any copyright or any other legal
right by Email communication. PHI will not accept any liability in
respect
of such communications.
signal command for shutdown - Howto?
Is there a how to out there to set up VM and Linux to respond to the 'CP SIGNAL SHUTDOWN USER' command? Our VM guy is suggesting that I make the console TTY just default to logged on. Had me make this change: >From /etc/inittab change this: # on S/390 enable console login in all runlevels #1:012356:respawn:/sbin/mingetty /dev/ttyS0 To this # experimental enable console login in all runlevels with /bin/bash -i 1:012356:respawn:/bin/bash -i /dev/ttyS0 This works, but I don't know that I like having a logged on console device just 'sitting there'. He wants to use it for some sort of automation widget within VM to shutdown or control the guests, but I'm not really comfortable with the exposure. Mebbey I'm paranoid but I think not in this case. -- 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: signal command not shutting down cleanly
On Mon, 25 Oct 2004 07:53:09 -0400, Levy, Alan <[EMAIL PROTECTED]> wrote: > When I tested out the signal on some of these machines, it took about 2 > minutes to shut them down. I think that 6 minutes is more than enough to > shut down linux machines but I am looking for further guidance on this. > Does anyone have similar experiences ? When the virtual machines are large and mostly paged out because they have been pretty idle, it takes quite some page faults to bring them in. If your z/VM DPA is overcommitted you'll find that the pages just brought in for one of the servers are dropped to make room for the others... I recall that bringing down 100 servers that way certainly took much longer than just bring down one of them... PS You can set the signal such that the FORCE command works like a SIGNAL SHUTDOWN. -- Rob van der Heij rvdheij @ gmail.com -- 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
signal command not shutting down cleanly
I have the operators issue a signal command to shut down all linux machines cleanly (CP SIGNAL SHUTDOWN USER xxx WITHIN 360). We IPL'd on Sunday and when I looked at the consoles this morning, all of the machines (about 10 of them) did not shut down cleanly. The signal command is working since I am seeing the shutdown in the VM console. Sometimes I see the following messages and sometimes I see that it gets hung up stopping Tomcat or Apache. Running /etc/init.d/halt.local"" Sending all processes the TERM signal ..done md: recovery thread got woken up ... md: recovery thread finished ... When I tested out the signal on some of these machines, it took about 2 minutes to shut them down. I think that 6 minutes is more than enough to shut down linux machines but I am looking for further guidance on this. Does anyone have similar experiences ? Alan Levy -- 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: Setting up VMNFS
| Gordon Wolfe wrote (22Oct2004 10:23): The syntax of the mount command you use from linux. What do you do about hard-coding the passwords in the mount command? Gordon, for general information about VMNFS, see http://www.vm.ibm.com/NFS/index.html. For MOUNT we use MOUNTPW, that avoids coding the password on the MOUNT command. The mountpw C source file is supplied by IBM as part of TCP/IP for VM on TCPMAINT's 592 disk. Issue the mountpw command for a certain disk userid.vaddr: mountpw node:userid.vaddr,u=userid,p=password Then give the real MOUNT command for that disk, omitting your userid and password: mount -t nfs node:userid.vaddr,trans=yes /mnt Check with mount query (just the mount command without operands) and see that no authentication data is displayed. Cheers, Loek Sluijter. -- 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: Problem with VM and Linux for zSeries
On Mon, 25 Oct 2004 12:21:12 +0200, Aristarc Diez <[EMAIL PROTECTED]> wrote: > It seems a problem caused by Suse Security Checks. > Any idea? Sure. You have made a very large virtual machine that did not see enough data so far to actually touch all those MBs. The SuSE process that started reads in a lot of data and Linux starts to use the MBs and needs z/VM to back that virtual memory with paging space. For each GB of virtual machine size that you define, you must allocate a 2 GB page pack (to repeat the message from Christian in other words). Rob -- Rob van der Heij rvdheij @ gmail.com -- 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
Problem with VM and Linux for zSeries
It seems a problem caused by Suse Security Checks. Any idea? TIA ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: - Remitido por ARISTARC DIEZ REDORTA/COSTAISA con fecha 25/10/2004 12:19 - ARISTARC DIEZ REDORTA Para: [EMAIL PROTECTED] cc: 25/10/2004 12:08 Asunto: Problem with VM and Linux for zSeries Hi, we have a big problem with our Linux Servers under VM. Every Sunday night our VM shuts down telling : HCPPGT401I 90 percent of all pagins space is in use. HCPPGT400I All paging space is in use HCPPGT401I 90 percent of all spooling spaces is in use. And then shuts down. Have you any idea of what are our Linux Servers doing? TIA. ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: -- 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: Problem with VM and Linux for zSeries
Linux on 390 Port <[EMAIL PROTECTED]> wrote on 25.10.2004 12:08:13: > Every Sunday night our VM shuts down telling : > > HCPPGT401I 90 percent of all pagins space is in use. > HCPPGT400I All paging space is in use > HCPPGT401I 90 percent of all spooling spaces is in use. > > And then shuts down. > Have you any idea of what are our Linux Servers doing? Very likely the cron job of the Linux guest will start some housekeeping jobs. As cron is triggered by the clock they all start at the same time. As a workaround you can modify the time for the cron jobs, so that each Linux has a different start time. Nevertheless, the problem is your paging configuration. You should definitely increase your paging space. There should be enough space to page the whole virtual memory of all guests (some people say you should have twice the amount of paging pages than virtual memory). As far as I know, you should never use more than 50% of the paging space for performance issues. This prevents fragmentation of paging space and allows VM to write to the paging device in larger chunks. Hope this helps -- Mit freundlichen Grüßen / Best Regards / Un cordial saludo Christian Bornträger Linux Software Engineer zSeries Linux & Virtualization IBM Deutschland Entwicklung GmbH email: [EMAIL PROTECTED] Tel +49 7031 16 1975 -- 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
Problem with VM and Linux for zSeries
Hi, we have a big problem with our Linux Servers under VM. Every Sunday night our VM shuts down telling : HCPPGT401I 90 percent of all pagins space is in use. HCPPGT400I All paging space is in use HCPPGT401I 90 percent of all spooling spaces is in use. And then shuts down. Have you any idea of what are our Linux Servers doing? TIA. ... : : : : : Aristarc Diez : [EMAIL PROTECTED]: : : Programador de Sistemas : Tel.: +34.93.2536100 : : : COSTAISA: Ext.: 209 : : : Barcelona : FAX : +34.93.2057917 : : :.: :: -- 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: virtual hipersockets
You configuration looks : world - qdio osa card - vm/TCPIP - hypersocket - LINUX. Mark , Pieter and Alan already outlined that hipersocket is not a p2p (but this is probably not the problem in this case) and pointed to verify routing problem both from linux to the world and viceversa. I only suggest to verify the routing into the osa card. OSA cards route only registered addresses to their unit addresses, and unregistered addresses to PRIMARY ROUTER (if defined) . So : - define [EMAIL PROTECTED] as PRIROUTER (if none is already defined) - OR try to use vmTCPIP as a PROXYARP (not sure you will be able, vmTCPIP ifconfig sees the link as a point2point , but the profile definition is not for a p2p, it is for an hypersocket !). hoping I didn't understand the configuration, and it helps, Franco Mignogna , IBM Italy, IT specialist -- 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: CAN-2004-0887
Hi Greg, > Can't find any info on this but it is mentioned here: > http://www.ussg.iu.edu/hypermail/linux/kernel/0410.2/2264.html CAN-2004-0887 is a local root exploit specific to s390. The only affected distro is SLES9 and they have a security update in place. Do you need to know anything more specific ? blue skies, Martin Martin Schwidefsky Linux for zSeries Development & Services IBM Deutschland Entwicklung GmbH -- 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
