Re: export DISPLAY
Mark wrote: >If you're using the -X option on the SSH command, and >if the SSH server on the Linux/390 system has >"X11Forwarding yes" set, then you should not >need to do _anything_ with the DISPLAY environment >variable. It should already be set to something > like "localhost:10.0," which means that SSH is >doing its >job of tunneling the X traffic back over the encrypted >link. >So, do the "ssh -X" command, and then as soon as >you're logged in, do an > "echo $DISPLAY" to see >if it is set or not. If not, examine the >/etc/ssh/sshd_config file to see what value it has for >X11Forwarding. With /etc/ssh/sshd_config X11Fowarding yes, dislay is set to 10.0 and the java scripts work. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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
export DISPLAY
I'm learning now to put DB2 and WAS on, both java installs. Before issuing the setup commands I set up as follows: fm workstation ws-ip, I issue ssh -X s390-ip fm s390-ip, I issue export DISPLAY=ws-ip:0.0 The work station is an i386 SLES9 under KDE. When YaST2 is issued on the s390, it uses GUI on the WS. When I try the installs for DB2 and WAS, they fail, saying the export is not set up correctly. Does anyone know the magical combination to get GUI to work (other than use console install mode). Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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: I want to move SLES9 under z/VM to LPAR.
"Is above process right?" None of the volumes can be minidisk. They must be all native volumes to work under an LPAR. If any of the volumes are minidisk, you have to do a logical copy from the minidisk to the native volume (cp -a or tar2tar). Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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: Latest on PAV's?
Alan wrote: "The primary reason PAVs exist is to provide relief to fully-populated z/OS systems with 256 chpids and 64K devices." That assumes an even access density on 256 chpids and 64k devices. Real world experience suggests that 80% of your devices have low access densities and you don't care about PAV's, but 20% of devices have a real impact on your overall performance with higher access densities, generally getting higher queing. It also glosses over the fact that each PAV requires a subchannel address, so they do very little to relieve the 64K device limit. Access density is also why the cache on dasd subsystems is effective - even though they have to grow as more dasd is put behind the cache. If the access density was even throughout a shop, cache would not be effective. With the current DASD implementations where 80% or more of active data comes directly from cache, even over FICON channels, the bottleneck becomes the software queueing causing by the subchannel contention. One might even benefit from PAV's on the "smaller" 3390-3 if the access density is high enough - 2.3 GB index that doesn't fit in memory, for example. In a VM environment, one probably doesn't get much queuing on minidisk. What about Page Devices? Its going to be interesting to see who VM pages on a 15288 cylinder paging device. What about full pack minidisks needed to hold large data bases (Oracle, DB2 UDB) or other high density data? "we can solve it administratively instead of with technology." hrmph! At some point, moving large amounts of data becomes impractical. And the VM systems programmers ranks are getting thinner. There are two bottom lines to this discussion. 1) The VM people assume that the guest OS will handle the PAV's, and 2) the real answer is to measure and determine where your bottlenecks are. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Sports - Sign up for Fantasy Baseball. http://baseball.fantasysports.yahoo.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: PAV's revisited
>Begging for 3390-1 is probably not a good idea in that >environment. Our zOS staff, with whom we have to share a box, and whom have all the PAV they need, is voting for a size of 15288 cylinders (a mod 13.7) - that gives the maximum usage of the backend store or so they say. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Yahoo! Messenger Show us what our next emoticon should look like. Join the fun. http://www.advision.webevents.yahoo.com/emoticontest -- 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: Latest on PAV's?
What about PAV's for page/spool volumes? We're trying to decide what size shark volumes to use in a new VM installation. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.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
Latest on PAV's?
What's the latest support for PAV's for VM? To support Linux volumes under VM? Our MVS people want to jump from 3390-3 to 3390-"12" images. No problem for them, they have PAV's. But we have to share the shark. Our preference would be to go to 3390-"12" if there is PAV support, otherwise a smaller device size. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Personals - Better first dates. More second dates. http://personals.yahoo.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
Linux 2.4 to 2.6 Transition Guide
I don't know if many of you monitor the IBM eServer: Linux on zSeries page http://www-1.ibm.com/servers/eserver/zseries/os/linux/index.html There are more and more links to whitepapers, tips and hints, redbooks (and of course, marketing stuff). An interesting white paper is available there dated this month (March,2005): Linux 2.4 to 2.6 Transition Guide. Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/ -- 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: Migration from Linux LPAR to VM
Nahant, I have several Linux that IPL either under VM or as an LPAR. They are setup up to IPL correctly as an LPAR then I use attaches under vm to use the disk as full disk. However, my LPARs and VM are on the same OSA. The gotcha there would be defining the right portname for an OSD if the OSA changes. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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: IP Drops out and comes back after 5-15 minutes
Have you checked to see if your OSA card is getting any errors? I've seen this kind of thing happen with CRC errors on the card. ===== Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.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
running commands at boot (SuSE vs RedHat)
I've found that I can issue commands at boot in SuSE using /etc/init.d/boot.local. Does RedHAT EL3 have a similar mechanism? ===== Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.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
VM direcotry entry for POSIXINFO
I noticed that the VM DIRMAINT entry for my Linux guest has a entry for "POSIXINFO". I thought this was only for MVS. Does it serve a useful purpose for a Linux guest? = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.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: Strangeness with the date command
>"All [...] z/VM systems will be IPLed Monday morning >between 6:00am and 8:00am EST for the >time change." This is often an excuse to get a lot of other stuff done too, like periodic system maintenance, and system upgrades. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). ___ Do you Yahoo!? Express yourself with Y! Messenger! Free. Download now. http://messenger.yahoo.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: Disk Performance
What kind of disk are you using (ramac III, rva, shark?). They have considerably different performance profiles. On a single channel, where you have to read, then write, this would be going at 1/2 device speed as there is no channel overlap. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). ___ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.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: mkinitrd and SLES9
Chris, the messages you get are normal for mkinitrd. After mkinitrd is run, you have to do a zipl to point to the new initrd module. For SuSE, /etc/zipl.conf contains the the dasd=,, values for the dasd. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). ___ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.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: Stateless Linux project
The procedure has the following statement: [StatelessLinux] name=Stateless Linux baseurl=http://people.redhat.com/dmalcolm/stateless Gripe: Why does RedHat insist that we connect to their sites to get updates (up2date) and do builds. I happen to have a number of systems that do NOT connect to the internet because of security reasons and putting U1 and U3 on via iso were a pain. Why would I want a procedure to maintain my systems that requires connection to a RedHat website. Why can't they include what you need on the iso's and allow you just to obtain the ISO's from their website? = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail -- 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: SLES9
Rick, I did all my installs under VM, then ipl'd them in an LPAR. The modprobe for cpint is looking for a module that talks to VM. I wonder if there is an assumption that the build is under VM because cpint is not used in LPAR mode. Have you contacted SuSE on this? ===== Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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: SLES-9 install notes
Mike, During the beta for sles9, there was a tarfile that built the installfile structure for you. I noticed that it was missing in the RC edition. It set up the environment described in the notes (and you implemented). I wonder where that install file disappeared to? = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail -- 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: LVM thread
Steve, did you do a mkinitrd between updating /etc/modules.conf and zipl? On RHEL3, the mkinitrd puts the dasd addresses in the initrd file, then the zipl updates the pointer to the initrd file. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail -- 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: SLES9 and SCSI
Gordon: The only relevant entry in /etc/sysconfig/kernel is INITRD_MODULES="reiserfs". mkinitrd is using some other mechanism to determine the modules for SLES9 (hotplug/coldplug). Greg: SLES9 has completely redone how it does the determination of the modules. It's not really like SLES8 at all with the addition of a hotplug/coldplug and udev mechanisms. I've looked in the mkinitrd code, but its not clear to me what to do. Also, I have tried changing /etc/sysconfig/scsidev from SCSIDEV_ARGS="r -q" to "disable" and that doesn't seem to turn it off. Whatever mechanism its using is new to SLES9 = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail -- 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
IPv4 and IPv6 on same OSA-X card
I have two LPARs sharing an OSA-GBX card. The network connection to them is IPv4. However, for testing, I need to talk IPV6 between the two LPARs. Is this possible on the same card? Or will I need a second OSA card. How do I set Linux up to speak IPv6 (or pointers to a procedure). Its SuSE8, so its already IPv6 enabled. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Take Yahoo! Mail with you! Get it on your mobile phone. http://mobile.yahoo.com/maildemo -- 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: SLES9 and SCSI
Alan wrote: >Are the FCP adapters defined in the IOCDS? If so, take >them out. There are no SCSI adapters on the zSeries and no SCSI defined in the IOCDS. All dasd is ECKD. mkinitrd, though, insistes on adding the scsi modules when it makes initrd. ===== Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail -- 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
SLES9 and SCSI
Now that its been announced, has anyone worked with the goldmaster of SLSE9 yet? Question: If I have a zSeries without SCSI, why is the SCSI subsystem started? How do I turn the SCSI subsystem off? Its addes quite a bit of time to an LPAR IPL as it polls each device available to the LPAR. (For VM, the choice is small and having the SCSI driver active may or may not be a problem). = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail is new and improved - Check it out! http://promotions.yahoo.com/new_mail -- 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
WG: Performance of large file systems
>Aren't the DASDs virtual in the storage boxes?? Tracks >are living >anywhere!? That may be true for the RVA, which is a log structured array (paging) device that maps a page anywhere in the backend. On a shark, each LSS is a separate RAID 5 group and device addresses are assigned sequentially in the LSS, so if multiple devices are assigned to the same LSS, they are limited to a speed of the LSS. For example, for on an E20 shark, each LSS has a bandwidth of 40 mb/sec, but the total bandwidth is about 160 mb sec. If you put all your stripes in a single LSS, you will only get 40 mb/sec, total, maximum. The only way to get the full bandwidth is to put volumes in the stipes in different LSS. I don't know the figures for later version sharks. I'll find out. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Vote for the stars of Yahoo!'s next ad campaign! http://advision.webevents.yahoo.com/yahoo/votelifeengine/ -- 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
WG: Performance of large file systems
>But we have big performance problems on an 40GB fs (LVM, >10 Stripes on >M9, 4 phys Escon) with an mySql-DB on it. Reorgs and even >selects >sometimes takes a lot of time. (Same on SCSI-Raid is >double-performance) Two things: 1) A reorg does both a read and a write, so the activity of the 4 escon channels are at least double and a 40GB write is overwhelming the cache on a Shark (or almost any device), so you are running, probably, at back end speed for the shark. 2) people often fall into the trap of putting all their volumes in a single shark LSS because it is easy to administer - consecutive addresses and zipl dasd is just a range! However, this will reduce the performance of th backend to a single LSS. To get the best performance on a striped volume, each LSS must be in a different LSS or you may lose any advantage there is to striping. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail -- 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
MD-dev and barrier
I have some doc that refers to "MD devices" and boot option "barrier". What devices do they mean and what does barrier do? = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Mail is new and improved - Check it out! http://promotions.yahoo.com/new_mail -- 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: Is there is an editor
ex is the line editor mode of vi - basically the same commands - and sed is a batch editor you can use. They both come with the distros and even work on the HMC console in a pinch. = Jim Sibley "Computers are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail -- 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
SuSE vs Red Hat
IMHO, SuSE seems to be more zSeries friendly. Some differences: - SuSE SLES8 allows the hertz timer to be turned on or off, depending on whether your are under VM or not. RedHat RHEL3 does not - hence you may have a performance issue with RedHat under VM with "idle" machines. - Redhat does not really support Reiserfs, if that is your preference. It allows mounting Reiser, but it does not include the tools necessary to support the filesystem (mkfs, tune, and the like). = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.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
Initial LPAR install zSeries
Gerard wrote: "Has anyone done a LPAR install on zSeries and then migrated to a zVM infrastructure ?, any ideas on this ?. I would look at the LPAR install just so that we get a head start on the project , the zVM is part of the IBM bundle that will come a little later in the project." Early on I did tape installs until we got a small VM to play with. Since then, I have done all my installs under VM to real DASD for LPAR use. Basically, the tape installs were slow and cumbersome, especially of the tapes were in a different room than the console and if there were errors and you had to start over. Otherwise, they worked satisfactorily. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.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: Article on Groklaw on SCO's subpoena of FSF
What else do they have to do in Linden, Utah, especially since they stopped polygamy = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Domains Claim yours for only $14.70/year http://smallbusiness.promotions.yahoo.com/offer -- 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: ssh access denied
Through the grape vine, I received info that /etc/ssh/sshd_config that prohibits sending clear test passwords from the client in a new release. To make putty behave like it used to when working with the new release simply comment out the line: PasswordAuthentication no = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Movies - Buy advance tickets for 'Shrek 2' http://movies.yahoo.com/showtimes/movie?mid=1808405861 -- 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
ssh access denied
I can login in via ssh from a linux terminal using ssh, but when I try loggin in using ssh from putty, I get an access denied. Which paramenter in the ssh.config file would affect this? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Yahoo! Movies - Buy advance tickets for 'Shrek 2' http://movies.yahoo.com/showtimes/movie?mid=1808405861 -- 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: cpint error messages
I don't see the cpint_load command in the 2.1.0 cpint, just hcp, monstat, and mongen. Will the 2.2 version work with 31 and 64 bit. ===== Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso (The NSHO's expressed here represents no-one but myself). __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: DB2 UDB on SLES-8 s390x
Jae-hwa wrote: "On the SLES 8 64bit(s390x with SP3), we(with our db2 engineer) installed DB2 UDB V8.1.5(FP5)." DB2 UDB currently is only supported on 31 bit zSeries Linux in DB2 UDB V8.1.5(FP5). 64 bit is NOT supported. If you check the readme file that came with FP5, you'll see the omission of UDB (though DB2 Connect is supported. (Or is it "not see the omission ;-)". Most other platforms are supported and DB2 Connect is supported at 64bit on zSeries Linux. I thought it was coming out with FP5, myself, but it didn't. I'd check with IBM to see if its scheduled for FP7 = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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
cpint error messages
I have cpint-2.1.0-1.1.s390x.rpm on a 2.6.5 s390x kernel. When I enter hcp "q v dasd" I am getting the error message "Open: No such device or address" 'Sup? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Newbie question about minumum installation
Rod wrote: "If you ask 10 people how you should partition things then you'll get 11 answers. A lot of it depends on what you want to do." Certain vendors always reply, "It depends", ;-) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Central processor is looping when 64 bit linux is IPLed
If you look at the mounted volume, if it is 64 bit, you will see a directory /lib64. If it is 32 bit, this direcotry is not present. Also, your mkintrd provides a hint: (version 2.4.19-3suse-SMP) (s390x) The notation s390 is 31bit, s390x is 64bit. So, you have 64bit. The zipl and mkinitrd look o.k. Its even finding the dasd_eckd_mod module that is sometimes left out. I guess the insmod set the environment for mkinitrd correctly. After doing all this, I don't see why you are looping during IPL on zero PSW's. Have you tried a 31bit install (assuming you have the materials)? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Control Unit is busy ( in use ).
"UNIT=2105,UNITADD=((00,50)),CUNUMBR=301 IODEVICE UNIT=3380,ADDRESS=(815,10),CUNUMBR=301,PARTITION=LINUX " UNIT=3380?? Are you really using 3380 images or are you using 3390 images? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Central processor is looping when 64 bit linux is IPLed
target=/mnt/boot/zipl image=/mnt/boot/kernel/image ramdisk=/mnt/boot/initrd parameters="dasd=1b26-1b28 root=/dev/dasda1" The /mnt mount should not be there. The purpose of the chroot is to make /mnt the "root" device. Also, since /usr is a separate volume, you need a bit more. After IPL'in the cd, mount /dev/dasda1 /mnt/newroot mount /dev/dasdb1 /mnt/newroot/usr mount proc proc /mnt/newroot/proc chroot /mnt/newroot edit /etc/zipl.conf such that the /mnt is removed run mkinitrd zipl exit (to get out of chroot) sync halt Then boot with your new system. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Central processor is looping when 64 bit linux is IPL
Taraka wrote: >Resending the same query again. >I am installing 64 bit linux SuSE 8.0 in LPAR mode on >zSeries. >Completed all the Yast partitioning and package >selection. Even installation using Yast went thru fine >without any problems. >When i IPL the Linux LPAR with the root dasd address, >it throws the following hardware error message. >An error was detected in partition LINXLPAR. Central >processor (CP) 00 is looping due to switching >between program status words (PSWs) that are not >valid. The program status word(PSW) is > - ( 32 zeroes). >Any inputs ? The psw and looping would indicate that zeros have been loaded into the first 4K or memory. When the boot loader branches to memory, an a zero psw is loaded, an interrupt occurs, a zero interrupt handler psw is loaded which then causes the recursive looping (reloading of the same zero psw). How are you building your system? from tape or cd? If tape, what tape drives? I think that I'm one of the few that actually use LPARs, but I tend to build my LPAR systems under VM because that is better tested and most often used. I found tape build was combersome. I assume that you have a SuSE License - I would report the problem to SuSE ASAP. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: IBM Virtualization Engine
Scott wrote: "Is this simply an extension of LPAR technology, or is it z/VM like hipervisor in the firmware? Is there anything new here for z/Series? Again, z/OS is mentioned, z/VM is not." The "virtualization" is not the kind we're used in VMWARE or z/VM or LPARs where a platform runs many copies of the different OS's. What this appears to be is a "virtual" layer between the OS/hardware and the application. The applications develop to the "virtualization layer" specs and hardware/OS vendor writes the hardware/OS specific implementation. This seems to be parallel to the Linux implemenations on various platforms - there is a part of the kernel that is hardware specific and the rest of Linux is a "virtualized" set of services that specify how the hardware implementation should react. If you look at DB2 UDB, there is an attempt to run it on all platforms. It would be simpler if they had to develop to a single set of specs and then each OS implement them. I would suspect that other software vendors would like the same capabilities. It would free them from limiting their market to a single platform (does windoze come to mind?) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Adding Spaces to a filesystem
>I picked /opt because the installation of IBM?s HTTP >server, Websphere and some other products are consuming >a major portion of the file system. You should be able to move just the IBM stuff to another volume, then put a link in /opt. 1. mount a device as /mnt/vendor 2. copy the IBM directories to /mnt/vendor 3. "rename" the directories in /opt, i.e., mkdir save, "mv IBMDIR save" 4. in /opt, link to the directories. 5. delete the /opt/save stuff when you are comfortable with the move. For example, for DB2, there is a /opt/IBM directory. Move the /opt/IBM directory to /mnt/vendor. Then, in opt, ln -s /mnt/vendor/IBM IBM Then you can refresh or rebuild your respack with having to reinstall all the IBM stuff. Some of the vendors have heard the message that people might want to install their stuff in directories the distributions don't write over. Maybe someday, they all will. (This does not imply a warranty on any vendors part that this technique will always work, just a fervent hope that they will get the message if they haven't already). = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: Adding Spaces to a filesystem
"create a separate file system for directory /opt " Out of curiosity, why /opt and not /usr. Have you done a du on /opt and /usr and compare the sizes? It would seem to me that, normally, /usr is the best choice. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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
term "zLinux" copyrighted by whom?
The official Big Blue legal position on the term "zLinux" is that it is copyrighted by "another company" and one should use the terms "Linux for zSeries" or "Linux for s/390". Anyone know who the other company is? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover -- 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: _re:_udev
Thanks, Arnd. I reported my take on the matter to SuSE via bugzilla, so another country heard from... "Also, creating device nodes by dasd label is tricky, because you need a device node if you want to read the label..." I just create 96 device notes when I build the image - that seems to handle most situations I run into. Right now, I'm using a cdl volser with the subchannel embeded (LX5700) and a file system label with the subchannel embedded (/LX5700). I would prefer only have to use the volser, but them days is gone, I think (I know, having the subchannel address embeded in the labels causes problems, but here they fix more problems that they cause). = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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 Storage Tuning Whitepaper for Red Hat Enterprise Linux 3
Mark, do they have the same kind of thing for 2.6 kenel? I now bdflush and its parameters have been reworked. And some of the other parameters have changed or disappeared as well. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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: udev on 2.6.5
Neale, I guess I don't yet understand sysfs and all that. I have a 3390 device, say 5700, at /dev/dasdb, I would rather use the subchannel address. What would I have to set up to issue the command: mount -t reiserfs /dev/dasd/0.0.5700/disc/part1 /mnt/test ??? I get the message "mount: special device /dev/dasd/0.0.5700/part1 does not exist" /etc/udev/udev.rules seems to know about s/390 DASD and partitions. Do I have to invoke udev? What would the command look like? Or do I invoke hotplug? Where would I expect to find the device in /dev? There does not seem to be a /dev/dasd direcotry. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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: sles9x beta (was udev on 2.6.5)
>I must have missed that. Is it publicly available? Rich, you'll have to check with SuSE. I couldn't say. ===== Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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
udev on 2.6.5
I've been working with the SuSE SLES9 beta, which is a 2.6.5 kernel. The April 16, 2004, "Device Drivers, Features and Commands" for the IBM site talks about udev and how it can can provide alternate device names. Has anyone made it work for zSeries? Is there a sample of how to enter the command, udev? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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
OT: SCO a new wrinkle
A new wrinkle "The new company, Open Source Risk Management (OSRM), set up by lawyer and venture capitalist Daniel Egger, plans to offer standard product liability insurance to major Linux vendors, big corporate users and individual developers of Linux." "After a rigorous six-month process of examining the individual software files in the Linux kernel and tracing their origins, OSRM found no copyright infringement in kernel versions 2.4 and 2.6," Egger, OSRM's chairman, said in an interview. He was referring to the two most commonly used versions of the Linux operating system." http://story.news.yahoo.com/news?tmpl=story&cid=580&e=5&u=/nm/20040419/bs_nm/tech_linux_legal_dc = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25" http://photos.yahoo.com/ph/print_splash -- 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: FTEs per Server
>I know a place where management allocate staff driven by >the sysprog/MIPS ratio... Let's see, in the early 70's we had a model 65 with an MVS maintenace person, an MVS build person, a JES2 guy, a program products person, an MVS debugger, and VTAM person, and a network person. So, that was about 7 people for under a MIP. Does that mean that with a z990 of 5400 mips, we should have 37,800 people supporting the z990? What a thought. LOL! We have about 15-20 people supporting about 80 LPARs (VM,MVS,Linux) with about 10 VM images, 1,000 MVS guests, 50 Linux guests, and 60 standalone LPARs on a total of about 45,000 mips. Shouldn't our software staff be 315,000 people? ROTFL! And that doesn't intclude the FE staff (we have 3 for the 45,000 MIPS plust 30-50 TB of dasd). We had 4 for the 65 in the 1970's - someone had to babysit the core memory add-on we had. The mind boggles (though it doesn't take much in my case :-) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html -- 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: FTEs per Server
>Darned users. If we could set traps for them and >eliminate them as a source of disruption, things would >be a lot easier. :-) I say give them fake keyboards >and mice. There is one school of thought that holds that users are just overhead, anyway ;-) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html -- 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: logout script from vm
To be more specific, forcing a linux system may not allow it to sync the filesystem metadata and data cache copy with the disk copy, so there may be damage to the filesystems after a force. If you allow a graceful shutdown, the filesystems are synced, greatly reducing the exposure to file system corruption. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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
BEWARE: virus masquerading as "fm: marist" with WORM_BAGLE.GEN-1 in MSG.zip attachment.
I received this in the company mail. Beware: --- fm: [EMAIL PROTECTED] subject: Hey, dude, it's me ^_^ :P Argh, i don't like the plaintext :) ..btw, "48615" is a password for archive - The company's anti-virus software deleted the zip that was attached with this message: Msg.zip is removed from here because it contains a virus. Found virus WORM_BAGLE.GEN-1 in file Msg.zip The file is deleted. Company's antivirus detection system has identified a virus in an attachment to this e-mail. The attachment has been deleted. No further reporting or action is required on your part. THIS EMAIL IS NOW SAFE TO OPEN. Visit w3.ibm.com/virus for more information. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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
ssh and telnet not working after rhel3 u1 install.
A colleague tried to build a new system using the RHEL3 U1 cds. After the install, he cannot login remotely from either ssh or telnet. He can login to the local console. Ping works successfully from outside hosts and he can ping out. However, when he does a telnet he gets a "No route to host" and ssh gets a "port 22: No route to host". What would be stopping both telnet and ssh logins? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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
Sun enters the Linux fray against microsoft via Walmart
Sorta off topic (different platform) http://news.com.com/2100-7344-5182158.html = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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: Adding QDIO guest lan to server
Could it be a syntax problem you have: lcs-1,0x3ad0,0x3ad1,0,3 qeth-1,0x0e04,0xe05,0x0e06,0,0 lcs1 and qeth1 without the dash. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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 disk for linux (CORRECTION)
Ken Hall wrote: >Basically, you need an initrd if the boot-device drivers >aren't built into the kernel. (You get this choice >during the kernel configure process.) With the variety >of hardware on that platform, ifyou tried to include all >possible drivers in the actual kernel, it >would be huge. So they make the drivers modules, and >load the ones they need. It appears that the SuSE and RedHat distributions use initrd on all platforms as a packaging technique to avoid having too many changes between kernels. That way they can have a standard kernel configuration. They then handle all the various device driver issues for each platform with initrd. That way, people can change hardware configurations without having to recompile the kernel. SuSE (SLES8) uses the $DASD parameter generated during the zipl process to set is dasd devices for zSeries. You need only change /etc/zipl.conf and zipl if you are adding devices that are already in initrd. Redhat (RHEL AS 3) has a different approach. They ignore the dasd parameters in /etc/zipl.conf and user /etc/modules.conf for their dasd list. The mkinitrd process reads /etc/modules.conf to build the dasd list. So, with RedHat, you have update /etc/modules.conf, run mkinitrd, then zipl.conf (to capture the new disk location of initrd...img). This has the curious affect that if you use the cio_ignore list, you have to update both /etc/modules.conf to change volumes needed and /etc/zipl.conf to change volumes to ignore. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/ -- 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: Myth of the 1K blocksize on eckd - revisited
Rod wrte: >>However, I could not >>recommend ext2 as a filesystem because it is too >>easily damanaged (sic) >How so? The Linux file systems are heavily cached and allow "dirty" writes of the data - i.e., the data in memory is not immediately synced with the hardware copy. This is controlled by the bdflush parameters. As the available memory gets largely, the more data is cached, and the exposure is greater. In 31 bit system, you might have maybe 1.5 gb cached. In a 64 bit system, this can be even more. The delay to write that much data may be several minutes during a sync, so your exposure for dirty writes could be several minutes of data! Ext2 is non-journaled, so data on disk can easily be corrupted by such things as a kernel panic or hardware failure or other catastrophic event. The entire filesysem can be come corrupted if the meta data is not in sync. Though a crash may be less frequent than say an MS system, they do occur occasionally. EXT3 (favored by RedHat) and REISERFS (favored by SuSE) have the meta-data journaled immediately. These are becoming the industry standard, replacing ext2. There is a much lower chance of corruption. The latest release of both distributions insist on a journaled file system during the build for the root volume. You can control how much data is dirty and how it is synced by use of the bdflush values. Setting the nfract value to 0 essentially allows few dirty writes. sysctl -w vm.bdflush="0 500 0 0 500 1000 5 0 0" Use these parameters with care as they do affect system performance. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: SuSE SP3 Kernel panic: VFS: Unable to mount root fs on 5e:01 ( was Re: Myth of the 1K blocksize on e
Before the zipl, did you do a mkinitrd? The dasd drivers are in the initrd module and not in the kernel. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: Myth of the 1K blocksize on eckd - revisited
In reply to Ron: >The best skewed distribution could be your own, so we >may get the horsedead by this: "Given this 4K >filesystem, would I save space when I made it 1K >blocks?" Agreed. I looked at /etc, which has "a lot of small files" and the average files size was 5900 bytes - not a goot candidate for a 1k blocked file system. However, as a ROT (for what they're worth), I would stick with 4K blocks unless you have the time to study your filesystems in detail and are willing to use different blocksizes for different filesizes. I doubt that the space savings would be worth the effort. In reply to Arndt: >Of course one might just use ReiserFS to get both the >space advantage for small files (even better than >ext2/1kb)and the performance advantage of using 4kb >blocks. I choose ext2 as it is, in some sense, the simplest to study (no journals, etc). However, I could not recommend ext2 as a filesystem because it is too easily damanaged. In reality, I used either ext3 for RedHat and reiser for SuSE. However, if I have a filesystem that may be mounted to either RedHat or Suse, I use ext3 because of RedHat Reiser support issues. I also ran the same experiments on Reiser and comparing 4k blocks of Reiser vs. Ext2 at 4k blocks and the numbers were slightly in favor of reiser: case1 max file size 1024 bytes 4k ext2 filesystem could contain 547,583 files, about 267 MB 4k reiser filesystem could contain 579,199 files, about 282 MB case2 max file size 2048 bytes 4k ext2 filesystem could contain 519,912 files, about 507 MB 4k reiser filesystem could contain 574,401 files, about 560 MB case3 maxfile size 4096 bytes 4k ext2 filesystem could contain 578,782 files, about 1,129 MB 4k reiser filesystem could contain 574,362 files, about 1,121 MB = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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
Myth of the 1K blocksize on eckd - revisited
Sorry for pushing this into the ground, but I'm still bothered by the "it depends" response for using 1K blocks on eckd devices. In the various file systems I've seen on various platforms, 1k blocks are only useful for very small files with none of them very big. So, to quantify this, I did some simple experiments to see where the cutoff for 1K and 4K blocks may be. I noticed in my previous overly enthusiastic experiments I posted that the packs were "were full", but there were 46% percent of the blocks left. Mark was correct in pointing out that I was running out of inodes. If you try and fill in a pack with very small files, you may hit an inode restriction, leaving free blocks on the pack. The default number of inodes for ext2 at 1K blocks is 206,848; for 4K blocks it is 300960 for a 3390-3 eckd image formatted respectively for 1k and 4k blocks. So the number of inodes needs to be increased. For mke2fs, specify a very large value for -N such as 200. For 1K blocks, this will give you 1,654,784 inodes and for 4K blocks, 622,592 inodes. So I formatted two dasd, both cdl, one with 1K blocksize and and ext2 filesystem of 1k blocks, maximum number of inodes, and the other with 4k blocksize and an ext2 filesystem of 4k blocks, maximum number of inodes. Another thing I found out was that the -h parameter for the df command gives you the total size of the blocks allocated, not the actual file sizes. So df reports actual space used, not actual total file size. Using these commands, then, in some sense, is misleading when trying to analyze your file system distribution. I set up four even data set distributions: Number minimum maximum avg total caseoffiles bytes bytes bytes bytes 1 10240 1023511 523,766 2 20480 204710232,096,128 3 40960 409520478,386,560 4 81920 8191409533,550,336 and copied each case, one at a time, onto the the 1k filesystem and the 4k filesystem until space ran out. What I found was: case1 1k filesystem could contain 1,418,912 files, about 691 MB 4k filesystem could contain 547,583 files, about 267 MB case2 1k filesystem could contain 900,597 files, about 878 MB 4k filesystem could contain 519,912 files, about 507 MB case3 1k filesystem could contain 588,266 files, about 1,148 MB 4k filesystem could contain 578,782 files, about 1,129 MB case4 1k filesystem could contain 312,557 files, about 1,221 MB 4k filesystem could contain 383,572 files, about 1.498 MB Conclusion: - if your average file size is much less that 2048 bytes and very few files are over 4095 bytes, then the 1K blocksize may be preferable. - if your average file size is about 2048 bytes and very few files are over 4095 bytes, then its a doesn't make much difference whether you use 1k or 4k blocksizes. - if your average file size is much more that 2048 bytes and many files are over 4095 bytes, then the 4k blocksize may be preferable. - I'm sure other more skewed distributions would give different results. But, in general, I would say that the 4k blocksize is preferable in the majority of cases. The space lost to eckd overhead is probably worse than any advantages you might get with small blocksizes. So, that is my cut at what "it depends" means. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: DB2 UDB 64bit now available
Rick, it looks like zSeries 64bit UDB may be in the next fixpack. But DB2 Connect is definitely 64bit. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: DB2 UDB 64bit now available
>Yes, but the URL you sent was for DB2 UDB V8 for z/OS >and only seemed >to >have links to general Linux info. It didn't mention SP5 >as providing >64-bit support for DB2 UDB for Linux for zSeries. Sorry, Rick, for the confusion. What's supported for Linux is buried in the README for fix pack 5. It only mentions db2 connect for 64bit, not UDB. I seem to have made another wrong assumption. ftp://ftp.software.ibm.com/ps/products/db2/fixes/english-us/db2linux390v8/fixpak/FP5_MI00073/Release.Notes/release.txt = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: DB2 UDB 64bit now available
Rick wrote: >Pardon me for being thick, but how does this affect >Linux for zSeries. >>On Mon, 2004-03-29 at 11:04, Jim Sibley wrote: >> In an earlier thread, someone was asking for DB2 UDB >> for 64 bit. It is now available. zSeries, if I'm not mistaken, is a 64bit machine. Prior to this announcement, DB2 was only supported for 32bit machines. At Fix pack 5 level, IBM DB2 UDB now supports 64bit on linux platforms, including zSeries. One of the threads on this forum asked about 64bit support for db2 UDB for Linux zSeries. At the time, it wasn't supported or announced. Now it is. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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
DB2 UDB 64bit now available
In an earlier thread, someone was asking for DB2 UDB for 64 bit. It is now available. http://www-306.ibm.com/software/data/db2/zos/db2zosv8.html = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: 3390-3 Dasd Capacity on SLES8 ext3
The maximum capacity of a 3390-3 formatted for 4k blocks is 2.3 GB (4096 x 12/track x 15 tracks/cyl x 3339 cylinders). Formatted for 1k blocks, the maximum capacity is 1.4 GB. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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: Myth of the 1K blocksize on eckd (corrected)
I was extrapolating based on the fact that both the 1k and 4k filesystems filled at 54%. So, I've tried to up the inode sizes. I can specify a large number (> 2M) for the 1k and I get 1,654,784 inodes, but I only am able to get 622,592 inodes for the 4k blocks. (-N 2,000,000 -i 8192). Yet there are more blocks on the 4k side. Why can't I get more inodes on the 4k size? ===== Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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
Myth of the 1K blocksize on eckd (corrected)
To answer both Mark and Tom 1) Both the 1k and 4k block experiment topped out at 54% of the file system full, so, yes, you do run out of inodes sooner, but you have more "capacity" for inodes with 4k blocks. Additionally, chasing the inode chain from disk (not cache) would be quicker with 4k blocks because there would be 1/4 the number of i/o's. 2) No, you do not get more blocks with 1k blocksize because of the heavy eckd overhead. 1k 4k blocks 1,626,740 2,366,164 And fewer blocks are used to contain directory information in the 4k case. If you increase the number of inodes, you will still get more small files on a 4k eckd image. Tom is right in that, if some of your files are more than 4k, then 4k rapidly becomes the choice, even if you are not inihibited by eckd formatting. Even if /etc, which are "small files", most of the files are greater than 1k and and a goodly number are over 4k. I don't see how 1k blocks are reasonable on 3390 ecked devices under any circumstances. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- 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
Myth of the 1K blocksize on eckd (corrected)
Sorry for the finger slip; I do not like MS products! In formatting a 3390 eckd image, it is clearly better to use 4096 than 1024 byte blocks! I formatted two 3390-3 volumes, one with 1K blocks and one with 4K blocks. I then put ext2 filesystems of 1k and 4bk blocks respectively. (SuSE SLES8 SP3) Using files from 0 bytes to 1024 bytes, I could get 1.2GB and 300,947 files on the 4k block but I could only get .812 GB and 206,741 files on the 1k blocks! Since I used from 0 to 1024byte files, all "small", 4k blocks still got more and files on the back. If I had used even bigger files, I would infer that 4k blocks would look even better. Here is a side by side comparison: 1k cdl 4k cdl fmt -b 1024 -b 4096 ext2-b 1024 -b 4096 blocks 1,626,740 2,366,164 empty 1.6 gb 2.3GB "full" 0.812 GB1.2 GB files 206,741 300,947 The file systems gave out before all the space was used - each filesystem showed 54% used, but you could not allocate any new files. The difference is space usage is because, with blocks greater than 4096 on a 3390 eckd, more "room" on the disk image is taken up with eckd overhead than data. For blocks gt that about 4096, there is more data on the track than overhead. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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
myth of the 1K blocksize on eckd
Formatting a 3390 eckd image, it is clearly better to use 4096 than 1024 byte blocks! I formatted two 3390-3 volumes, one with 1K blocks and one with 4K blocks. I then put ext2 filesystems of 1k and 4bk blocks respectively. (SuSE SLES8 SP3) I could get 1.2GB and 300,947 files of 1023 bytes or less on the 4k block but I could only get .812 GB and 206,741 files on the 1k blocks! Since I used from 0 to 1024byte files, all "small", 4k blocks still got more and files on the back. If I had used even bigger files, I would assume that 4k blocks would look even better. Here is a side by side comparison: 1k cdl 4k cdl 1626740 = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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: Format disk and mount FS : lost space
Adam wrote: >Depends. 1K blocks will be slower but less wasteful of >space if you have lots of files smaller than 4K. For SCSI devices, I would agree with you. For eckd devices, the overhead on 1K blocks is tremendous so what you make up from "small" files is lost to eckd overhead. Also, for eckd devices, the transfer overhead of moving 1k blocks is greater than the moving 4k blocks. I would suggest that, as long as you have eckd devices, 4k is better than 1k formatted devices. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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: DB2 Connect for 64 bit Linux - an even better reference
http://www-306.ibm.com/software/data/db2/linux/validate/rhel3.html = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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: DB2 Connect for 64 bit Linux
I suspect you'll see something about db2 connect and 64bit at the end of March. http://www.infoworld.com/article/04/03/03/HNmainframedb2_1.html ===== Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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: Certifying prodcuts on linux over multiple platorms
Dave wrote: >AFS tracks the OS releases pretty well. For the >last two major AFS releases, it's been about a week lag >at worst, if notb available first day (depending on how >cooperative the vendors have been on granting access to >new platforms and OSes). Therein lies the rub. We start some of the development during the Linux Beta Cycle, where NFS is reasonably stable, but AFS has not quite up to speed. For products later in the development cycle, AFS is certainly a good alternative. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.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
Certifying prodcuts on linux over multiple platorms
(Longish, Friday afternoon musings). We are running into some interesting problems in testing Linux across platforms. It is the opposite of building penguin colonies or rookeries or rafts or waddles. The scenario is common enough. We have a database product that goes through release and fix cycles. We then have several products that depend on the database product with their own release and fix cycles dependent on the database product. When we send out release 487 of the db, then all other products have to support release 487. Nothing really new hear. Linux has added a new dimension to this with the multiplicity of distributions and releases and the multiplicity of platforms that support it. We develop it say on intel and it seems to work. Does it work on all other platforms and release levels we day it does? For example, just considering enterprise releases, we have RH 7.1, 7.2, RH AS 2, RH AS3, RH AS3 U1, SLES7, SLES7 SP1, SLES8, SLES8 SP3, the uncommon thread being different vendor configurations, different kernel levels, and different maintenance levels, some or all of which run on our beloved dinosaurs, power PC's, intel, itanium, Sun, and HP platforms (just the ones we support now). So what does the test matrix look like? A subset might look like, with 23 different hosts required to complete the testing. intel intel ppc ppc zseries zeries 31bit 64 31 64 31 64 RH7.1 X RH7.2 X RH AS2 X RH AS3 X X X X X X SLES7 X X SLES8 X X X X X X SLES8-3 X X X X X X And, obviously, the developers, testers, and users want a common environment; they don't want to change anything as they move from host to host. They want to repeat the same certification on each host. The question, then becomes, what do I have to do to set up this environment? Some thoughts come to mind: 1 - each group needs a common storage for their work. NFS appears to be the best choice because it tracks distributions better than say AFS, which may not work with bleeding edge products. 2 - With a common storage, userids and groups (uid/gid) need to be consistent amongst all platforms to access the data. 3 - The build of each host has to be essentially the same as to products installed and not installed. (a good example of this requirement is the compatibility rpm dragged along from earlier distributions) 4 - There is an install test phase, where the installation process needs to be certfied and this invariably requires root access and an isolated system host. 5 - There is a function test phase, where all the users could conceivably use the same host, but we would need hosts at each level. So, does anyone have any thoughts on how to approach this problem? Any references you know of? (I don't like the word "challenge", I prefer the old mathematical connotation of "problem"). ;-) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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
OT: Penguin terminology, was RE: VSWITCH and LINUX
"According to Webster, a "rookery" is "The breeding place of a colony of," so colony would be the collection of birds, rookery is where they breed. So, when is IBM going to come out with a new version of the zSeries boxes code-named "rookery?" :)" So much for relying on the advanced education of my grandchildren ;-) Would "colony" and "ROOKery" be acceptable terms to a multinational company, considering the other connotations of the words? Maybe we should have a meeting? ;-) The company, though, seems to prefer dinosaur names for the big iron (raptor, trexx). = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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
Lindows (oops, sorry) LindowsOS
http://story.news.yahoo.com/news?tmpl=story&cid=581&e=1&u=/nm/20040312/tc_nm/tech_lindows_dc = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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
VSWITCH and LINUX
"Sorry should it be Flock, Herd, or School for penguins." The correct term seems to be "rookery" of penguins. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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: Can't locate module char-major-10-224
Malcom, this seem to a problem introduced by SuSE at the SLES8 SP3 problem. I am seeing the same thing on an unmodified SLES8 SP3, both 31bit and 64bit. As a temporary workaround, I have put the following in /etc/modules.conf: alias char-major-10-224 off It turns the message off and I am seen no ill effects - YMMV. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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: telnet limiting logins to 1 or 2
Mark, thanks for the hint - xinetd.log messages indicates that I have to increase the instances limit in /etc/xinetd.conf or remove it. defaults { ... instances = 10 } I was searching for telnet. When I searched for the error message on Google I found one hit - in the Man pages for xinetd.conf! = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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: telnet limiting logins to 1 or 2
Mark, the /var/log/xinetd.log shows FAIL: telnet service_limit from (ip) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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
telnet limiting logins to 1 or 2
I have telnet running under xinetd (SuSE SLES8 SP3). However, after the first user or two is logged in, subsequent logins get the message "Connection closed by foreign host." How do I get more users on with telnet? Am I missing a config file for telnet? Is PAM getting in the way? (Yes, I know they should use ssh, but there are some technical reasons why they cannot). = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.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: automatically starting xinetd in SuSE SLES8
Holger wrote: >use yast (network service configuration) or >create the appropriate symlinks in /etc/rc.d >manually Yast insists on starting inetd, not xinetd. The chkconfig puts the right entries in init.d to start xinetd. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.com
Re: automatically starting xinetd in SuSE SLES8
Duh! Thanks! And its not even Friday! = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.com
automatically starting xinetd in SuSE SLES8
Sorry for the brain fog. How does one start xinetd automicatilly in SuSE SLES8 AFTER the intial install? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Search - Find what you re looking for faster http://search.yahoo.com
Re: Linux and OSA/e Ethernet Questions
"How can I tell that these OSA connections are running at 100 and not 10 mb?" On the hardware SE or HMC, you can go into single object mode and look at the CHPs (you need to know the CHIPD of the OSA card). There is an icon for Advanced functions, which shows you the state of the OSA CHP and you can "view" the settings. They should show 100mb and full duplex. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Get better spam protection with Yahoo! Mail. http://antispam.yahoo.com/tools
Re: z/VM access to EMC (was: Accessing DASD on a Shark from Linux under z/VM)
>2GByte/sec is the speed of a single STI bus and the >smallest T-Rex (one book) has 12 STI buses while the >largest (four books) has 48 STI buses for a total of >96GByte/sec bandwidth. Malcolm, I stand corrected. The TREXX is truely an impressive I/O engine. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Get better spam protection with Yahoo! Mail. http://antispam.yahoo.com/tools
Re: z/VM access to EMC (was: Accessing DASD on a Shark from Linux under z/VM)
> > I'm curious. One of the benefits touted, and true, about > Linux on zSeries > > vs. some other platform, is the zSeries' strength in I/O. > Is this still true > > with FCP attached SCSI DASD? Why would the zSeries drive > SCSI DASD better > > than Intel or Sun? > > John McKown > > Senior Systems Programmer Basically you can attach more dasd space and have more simultaneous (NOT just concurrent) data transfers going on at the same time. The I/O advantage of the mainframe is that it usually has more paths (256 channels) to more devices(65,536) thus giving a lot more parallel I/O, not that any particular device is more efficient. If you have a lot threads active, more I/O can be done in parallel that most intel and other boxes. With 256 channels at say 12 MB/sec (shark) on , the total aggregate rate of the mainframe would be about 3 GB/sec. Obviously, that's limited by the 2 GB backend buss on the TREXX. Also, the theoretical limit with 65,536 devices, each with 32GB, would be something like 2 exobytes! (1024x1024xGB) (correct if I used the wrong prefix - its bigger than terabyte and petabyte). Adding SCSI/Fibre channel, the rates also increase, but the number of maximum channels is reduced. Also, the main frame typically has 2 processors dedicated to driving the devices (SAPs), so less "real cpu" is used for I/O. The other processors can continue to do other work while the SAPs are doing data transfer to the devices. Only after the data transfer is complete need the "real" processors be notified. Path selection, etc,etc, has been moved to these SAPS. The mainframes make dandy servers when you need access to a lot of data quickly. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Get better spam protection with Yahoo! Mail. http://antispam.yahoo.com/tools
Re: RHEL3 for s390x gcc 2.95 compatibility libraries
It looks like the following is the one your're looking for. Try rpm -q -p -i compat-libstdc++-7.2-2.95.3.77.s390x.rpm ===== Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - VSAM tangent
John wrote: "Oh, nevermind... no need to educate me about this, especially on this fine list. I don't currently do mainframes for a living, and z/Linux really has nothing to do with z/OS's Media Manager (as far as I know)." The one thing of interest is that media manager is also the driver for HFS/ZFS for Unix Systems Services on MVS. Its interesting to those that compare Linux on zSeries vs USS. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - VSAM tangent
John wrote: >I wouldn't say that VSAM is standardized at 4K blocks. >In general: >A Control Interval can be any size from 512 to 8192 >bytes in increments >of 512 bytes, and from 8 KB to 32 KB in increments of >2 KB. >The underlying physical record size is chosen by VSAM, >and depends on device geometry. The physical record >size is equal to or less than the CI size, and divides >into CI size with no remainder. For quite some time, VSAM has been using the media manager as its device driver. The standard block on disk is 4K and all I/O's are 4K, even though the CI may change in size. Unless you have something older than a 3380 or an very old MVS, a track dump would show that the actual blocks on disk are 4K. Also, if you looked at PDSE data sets and QSAM extended data sets, the blocksize on disk will be 4K! MVS has for some time been moving towards a fixed 4K block. The problem is the data in the field has to catch up. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - still not recommde d?
Actually, MVS benefitted from FBA devices, too. The resistance is having a mix of devices to handle FBA and ECKD. "If this is true, then why doesn't some enterprizing vendor sell disk that emulates FBA devices, to be used by Linux, VSE and VM?" "enterprizing" is an interesting word. SCSI is now available on the Enterprise Storage Server (Shark) and I would assume FBA (thougs it won't fill the bill for VSE or VM). I know we have a number of AIX boxes and intel boxes connected to Shark, but so far I have not had the opportunity to try SCSI on zSeries. Perhaps Neale could speak to the space usage of disk on SCSI for zSeries, formatted block sizes, and the like... = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - still not recommde d?
Rick wrote: "Another datapoint indicating that MVS should learn FBA. But it's a losing battle. There seems to be some unwritten law that says "disk" on "channel" must be [E]CKD, so that while VM, VSE, Linux, and others are happily embracing fixed block disk, the mainframe channel-attached "disk" MARKET stays constrained." Actually, it is a well know economic law. If people don't buy it, why support it? When os/370 came out, it introduced a fixed block file system (VSAM) now standardized at 4K blocks, eerily similar to the block size of a page in memory and the page size on disk. Also introduced were fixed block devices that would handle both VSAM and paging nicely. Few people bought them because most of their data were in odd size blocks, mostly multiples of 80 and they didn't want to have a mix of devices - some for VSAM and paging and some for the rest of their data. And to this day, a good portion of the data of the people who buy zSeries day are bound by the 80 byte record and variable sized blocks. Its pretty much the same reason people buy 2x4's (boards). That's a standard and its been around a very long time, so housing and building construction is based on a standard wood 2x4, even if the material is steel! A hollow steel 2x4 is stronger than a solid wood 2x4, but the measurement is still a 2x4! = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - still not recommend?
I just did a quick experiment: 3390-3 disk image formatted at 1K blocks, ext2 fs 1K blocks. df showed 1.6 GB capactiy 3390-3 disk image formatted at 4K blocks, ext2 fs 4k blocks. df showed 2.3 GB capacity. The drop in total device capacity was about 27%. I was not able to mount a disk formatted at 4k blocks with an ext2 fs of 1k blocks or an reiser fs of 1k blocks. Has anyone been able to to this? The reason is there is so much drop in capacity is that IBM settled on the 3390 image with a high overhead for small blocks. Even with the raid devices (RAMAC, RVA, SHARK, EMC, etc), the track is recomposed into a full eckd image in the cache and I/O is from the cache image. 4K makes a lot of sense when your paging is 4k blocks and most of your data is VSAM (4K blocks). As long as you have to use a 3390 track image with ECKD mapping, the smaller tracksize will cost you capacity on the image. Probably only the RVA will see equivalent savings on the backend. Other devices do a one to one mapping of the full ECKD image. FBA would be a better solution. Does the new SCSI support for shark allow 1k block sizes? I can see some other issues with 1k blocks. What would happen if you make the swap (page) packs 1K? Would it even work. With files over 4k, there would also be a loss in I/O efficiency - 4 times as much I/O would be required. And this I/O efficiency was also translate to reduced efficiencies for reading the node tables. With smaller node tables, there may be more I/O activity locating a file. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Re: dasdfmt with a 1K block size - still not recommded?
Beware of the small blocks on ECKD images. In a 3390 image, there are 12 4K blocks/track or 49152 bytes. There are 33 1K blocks/track or 33792 bytes! You could lose up to 1/3 of your capacity using smaller block sizes! = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
products on SuSE SLES8
Is anyone running any of the following products on SuSE SLES8 or SLES8+SP3 for zSeries? Java Secure Soecket Extension (JSSE) 103 Install Shield Professional MultiPlatform 5.02 AFS XSD Eclipse and EMF Version 2.0.0 Opend Development Environment (ODE) Install Sharity 2.8 Any info would be helpful. Thanks... = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance: Get your refund fast by filing online. http://taxes.yahoo.com/filing.html
Re: How much $$
"I would think not. The MP effect usually guarantees that 1 2-processor system is significantly less than 2 1-processor systems. The amount of lost capacity has decreased over the years, but it is still non-zero." 2 processors, running at capacity, would be about 1.8x of 1 processor. However, few people run at capacity. Even at low loads, there is an advantge to SMP because Linux is multi-threaded and I/O is asynchronous. SMP would should even help striped LVM. Even compiles can be speeded up if you can use the -j option (-j n+1, where n is the number of available processors > 1. The n+1 allows cpu and I/O overlap). The more concurrent activity, the better off you are with SMP, even with the "loss" due to spin locking. = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance: Get your refund fast by filing online. http://taxes.yahoo.com/filing.html
Re: zipl.conf for RH 3.0
I just did a quick experiment with RHEL3, mkinitrd, and zipl. 1) I updated /etc/modules.conf, then did the mkinitrd using the same name for /boot/initrd When I rebooted, the new dasd were NOT there. 2) I then did a zipl and rebooted. The new dasd was picked up. Bottom line - both the mkinitrd and zipl are required to change the dasd addresses. The "dasd=" is not placed in the boot record by zipl or mkinitrd. Is it that the boot record has to be updated by zipl to get to the new initrd module? = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance: Get your refund fast by filing online. http://taxes.yahoo.com/filing.html
Re: zipl.conf for RH 3.0
>The mkinitrd makes the /etc/modules.conf change > "permanent" in the /boot/initrd... module. >Then running zipl will make it permanent in the boot >sector of the IPL >disk. If you use the same initrd name, then the zipl may not be necessary. I think the zipl is only needed if the /etc/zipl.conf changes because the initrd name changes. (I'll verify this). To use the same initrd name, use this form of the mkinitrd: mkinitrd -f /boot/initrd-($uname -r).img $(uname -r) = Jim Sibley RHCT, Implementor of Linux on zSeries "Computer are useless.They can only give answers." Pablo Picasso __ Do you Yahoo!? Yahoo! Finance: Get your refund fast by filing online. http://taxes.yahoo.com/filing.html