Re: Linux DCS and swap

2.4.x you would either need to activate block device and swap manually after system startup or write your own /etc/init.d script that does that. You need at least Sles8 SP3+ (submarine). with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum h

Re: DCSS updates in Bitkeeper

Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the

DCSS updates in Bitkeeper

be applied against 2.6.10 anymore, it needs to be adapted to the new (fixed) extmem interface with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -

Re: Question re: Linux SLES9

ou can change that value by typing echo "1" >0.0.1234/online And using cat as above, you can check if that was successfull. This way of configuration works for everything that has a single device number (dasd, tape, ctc etc.) but not for those devices with multiple device number

Re: cmsfs on 2.6?

Should work fine with a loopback device with 512byte/block. Indeed I think it would be really nice to have such filesystem in the vanilla kernel series cause one can do very nice things with it. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, non

Re: Big Iron change

ns for the target platform. As for the kernel from distribution partners, I do not know about any middleware product that has done this. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo haben

Re: Big Iron change

800/900 series. No need to reinstall. You will still see the performance advantage of the new engines. Note that gcc can optimize for 890/990 too, the result would be a little faster but it would not run on 800/900 anymore. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit,

Re: DASD Usage

ow much space of the physical volume is associated with logical volumes. Whether or not the fragments of a logical volume found on the physical one do contain user data cannot be figured out with pvdisplay as far as I can see. with kind regards Carsten Otte -- omnis enim res, quae dando non defi

Re: DASD Usage

olumes based on your storage requirements instead of creating them based on the size of your hard drives. This makes it irrelevant which data is stored where, therefore I see no need to check how much free disk space is left on which physical volume. Why do you want to figure that? with kind regar

Re: Linux on Mainframes

We support G5 and up. You can also run it -unsupported- on older machines without IEEE fpu but you need to enable the FPU emulation option on kernel build - and the performance will be very limited on math opertaions. You should at least have a G3 machine for this... with kind regards Carsten

Re: IFL questions

ou please post the contents/output of the following files/commands at a time when you see the problem? > cat /proc/loadavg > pstree > vmstat 10 6 (samples 1 minute) PS: please put me on CC: when posting with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habet

Re: Running servers on z/VM vs on Linux

h kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED

Re: Shared /usr across linux guests - what about /var ?

This causes to rpm writing the parts on the wrieable non-shared filesystem while not aborting becuase of write errors on the read-only shared parts. We aim at solving this in the future... with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum ha

Re: Problems with SLES 9 and guest lan

agically instead of complaining about missing symbols. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / ar

Re: CP DEF STOR CONFIG

it detected). with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / archive access instructions, send

Re: Diag and FBA in 2.6

le parameter like this: modprobe dasd_mod "dasd=1234,1235(diag,ro),1236-1240,1241-1250(diag)" - and please do not forget to quote it because the shell would mess with the braces... with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetu

Re: Diag and FBA in 2.6

ameter, and in the end set it online again. Kernel should display a message to make it clear ;). with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est ---

Re: VDISK swap on SLES9 and the FBA vs. DIAG driver

le to me that a patch that makes the system crash on some configurations should be called "fix". Anyway, you can use the parameter (diag) in your parmfile/kernel parameter list to choose diag for a specific device or range. Works like this: "dasd=1234,1235(diag,ro),1236-1240,1241-1

Re: linux swap to dcss?

as our measurements we have so far indicate a very strong advantage in using DCSS for swapping over VDISK. Go pick up the code from kernel.org and measure it, or just stop complaining. By-the-way: Where is your proof for your statement? with kind regards Carsten Otte -- omnis enim res, quae dando n

Re: linux swap to dcss?

people are much better off when I keep developing good soloutions that integrate Linux and VM instead of trying to prove my points by performance measurements. Maybe you should visit one of either Rob's or my presentations to learn about current developmen

Re: linux swap to dcss?

>In the case of swapping to EW DCSS it would be a separate address space so the rules are not the same as >the XIP2 type of mapped file system. >David Nope, you can do the same with xip2fs. You can create storage holes and fill them with your filesystem. with kind regards Carsten Otte

Re: linux swap to dcss?

t in a customer distribution, but you can use a Vanilla 2.6. kernel from kernel.org to play with it. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -

[no subject]

re pushing the same idea here. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / archive access instruct

Re: linux swap to dcss?

swap, you can use 2gig dcss high prio, and 498gig on dasds. The VM paging priority issue you describe does not apply to DCSS based swap. What's your point? with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habend

Re: linux nss

27;s what autosave means. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est -- For LINUX-390 subscribe / signoff / archive access instructio

[no subject]

On Nov 17, 2004, at 10:32 AM, Carsten Otte wrote: >> You might as well be interrested in our DCSS execute-in-place >> filesystem. >> This one allows you to put your >> applications and shared libraries in DCSS segments, and they can be >> used/executed on individual

[no subject]

e-in-place filesystem. This one allows you to put your applications and shared libraries in DCSS segments, and they can be used/executed on individual guests without copying into virtual guest storage. with kind regards Carsten Otte with kind regards Carsten Otte -- omnis enim res, quae dando n

Re: linux nss

quite horrible. We are looking forward to do a better implementation with the new kbuild system that allows to set kernel parameters at IPL time via "parm", to have to kernel autosave itself into an NSS, and to with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum

Re: subversion

Neale, what exactly was the problem with rebuiling the 386 source package? with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est "Ferguson, Neale" <[EMAIL PROTECTED]> Sent by: Linux on 390 Port &

Re: Flushing the linux buffers periodically - pitfalls?

of similar servers - use collaborative memory management to synchronize the memory managers In effect, above soloutions help to keep your Linux+VM happy (and therefore keep your application performance up) even when memory pressure in Linux or VM gets high. with kind regards Carsten Otte -- omnis

Re: Determining what tasks own pages in swap?

ample). If your swap-in and swap-out rates are low during the workday, you do not have a problem at all. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est James Melin <[EMAIL PROTECTED]> Sent by: Linux on

Re: Best Practices EVMS FCP Multipathing

Hi Cameron, you definetly want to do multipathing with FCP. Unlike with FICON, your adapter (and your physical fibre) will be a single point of failure if you don't. In order to maximize performance, you can combine multipathing with raid0 striping to gain further. with kind regards Carsten

Re: how to recompiling linux kernel

option] - z890/z990 [only runs on z890/z990 only and is slightly faster than running with above option] When compiling a 64bit kernel, you only have the last 2 options. with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, q

Re: Shared filesystems with SuSE SLES 8

how to set up sharing with dcss: [for linux-2.4 on SLES8] http://oss.software.ibm.com/linux390/docu/lx24dcss01.pdf [for linux-2.6 on SLES9] http://oss.software.ibm.com/linux390/docu/lx26dcss00.pdf with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum ha

Re: Cloneing Linux Guests on FCP SCSI

ereadpt /dev/sdy - sync disks $bash> sync --> done with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est "Seader, Cameron" <[EMAIL PROTECTED]> Sent by: Linux on 390 Port <[EMAIL PROTECTED]>

Re: Cloning Linux systems problem

roc/dasd/devices - detach the dasd - attach it back to the guest that should run it - ipl just FYI: with sles9 you can edit the parameters at ipl time, that makes things much easier with kind regards Carsten Otte -- omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetu

Re: VM Shutdown

d that parameter line. On the other hand you can use "parameters=" to specify the parameters in zipl.conf and the tool will create a parmfile with the content for you. (*) this sentence was shamelessly stolen from Bill Bitner with kind regards Carsten Otte -- omnis enim res, quae da

Re: z800 performance

IO, CPU utilization may not become a problem afaics. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world -- For LINUX-390 subscribe

Re: Model 27 dasd

. In case you still run into a problem, please let me know. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world "Levy, Alan" <[EMAIL PROTECTED]> Sent by: Linux on 390 Port <[EMAIL P

Re: raid1

--Original Message- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Carsten Otte Sent: Freitag, 13. August 2004 10:21 To: [EMAIL PROTECTED] Subject: Re: raid1 Hi Martin, you cannot use a raid1 device for boot itself. The zipl bootloader requires either a scsi disk or a dasd.

Re: raid1

update your system kernel. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world Lufthansa Systems Infratec ZLINUX SYSPROG <[EMAIL PROTECTED]> Sent by: Linux on 390 Port <[EMAIL PROTECTED]> 13/0

Re: Linux 390 tape drive

Hi, have a look at /proc/tapedevices. This file is available as soon as you loaded your tape390 device driver. It tells you what devno's are currently managed by the driver, and what device nodes correspond with them. with kind regards Carsten Otte -- I saw screens of green, red message

Re: 2.6 kernels and NSSes

we do not support it. Please be patient for a while, we're planning to work on a nicer way to get the kernel into NSS... with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world Adam Thornton <[EMAIL PR

Re: DASD Errors due to R/O

our parmfile or when using /proc/dasd/devices to add devices you can use (ro) after the device number to preset the device to be read-only. If anyone knows a good method to autodetect if a device is read-only or not, please let us know! with kind regards Carsten Otte -- I saw screens of green, red

RE : Kernel memory allocation

se significantly but you still see the slowdown, another idea is to look at VM to see if it is memory constrained. If this is the case, you need to _decrease_ the size of this and other virtual machines in order to fix the problem. This is a balancing issue. with kind regards Carsten Otte -- I s

Re: dasd_diag_mod mission in SLES8 64bit

sion of the diagnose. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful wo -- For LINUX-390 subscribe / signoff / archive access instructions,

Re: Myth of the 1K blocksize on eckd (corrected)

case of sufficient inode space you'll be able to store far more small files (1..1024 bytes each) on a 1k formatted dasd. The filesystems will be 100% full in case you have sufficient amount of inodes. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue,

Re: Format disk and mount FS : lost space

not like that overhead can just use LDL which only uses 2 blocks (8 kilobytes) as overhead. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonde

Re: filesystem overhead

.. is there any reason to favor that method over "cp -a" on >systems that support it? Afaics cp -a should do just the same thing. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to my

Re: Using DCSS for extra memory

le (example in 1meg steps), you would need a lot of DCSSes because those are so far not relocateable and cannot be mapped multiple times by the same guest. These restrictions make the "DCSS extends memory" idea unattractive. with kind regards Carsten Otte -- I saw screens of green, red messa

Re: bk 2.6 & s/390

Hi Geoff, just use the vanilla tree. S390 is well integrated, consider reporting bugs to us in case you find some! with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world "Geoff O'Callagha

Re: New xip2fs patches on the web site

ems best to share executeable files and libraries that are frequently used by a large amount of Linux images, and it is counterproductive to share things that are not frequently used. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think t

Re: New xip2fs patches on the web site

Hi Alan, as far as I know, ROMfs does not support execute in place? What is done here is that userspace adress tables effectively refer into the shared memory segment. (see nopage and mmap functions in fs/xip2fs/file.c) with kind regards Carsten Otte -- I saw screens of green, red messages too

Re: New xip2fs patches on the web site

filesystem on it, and put executable files and libraries onto the filesystem - save the dcss segment using the block device - reboot and issue a command like mount -t xip2 -o ro,memarea=MYDCSS none /mnt to mount the new filesystem and gain memory resources 8-)) with kind regards Carsten Otte

Re: dasdfmt with a 1K block size - still not recommded?

and Linux. This leads to the general recommendation to use 4096 block size because it provides best efficiency and performance with a typical (average) file size greater then 2048. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to mys

Re: DCSS filesystem setup script

/sbin, /lib, etc. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world Adam Thornton <[EMAIL PROTECTED]> Sent by: Linux on 390 Port <[EMAIL PROTECTED]> 17/02/2004 11:27 PM Please respond to

Re: DCSS permission question

in /proc/mounts. (try to issue "mount" without arguments versus "cat /proc/mounts" to see the effekt). with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world Adam Thornton <

Re: DCSS permission question

>I see greater use of FS labels coming from this. >The driver will need to work with /proc/partitions >because that is where 'mount' reads filesystem labels. Why would you want to use FS labels? Just make your DCSS have talkative names! with kind regards Carsten Otte -- I s

Re: Please help with DCSS

Hi Adam! You need to save a segment, you now have a skelleton (CL "S") in NSS Map. Do #cp def store 256M, then #cp saveseg SHRUSR. Now you should have it active (CL "S"). With guest storage <=128M and mem=256M parameter it should work fine. with kind regards Carsten O

Re: vm guest waitstate

; or "Oops" followed by some hexadecimal register and stack dumps. with kind regards Carsten Otte -- I saw screens of green, red messages too, then came blue, shubidu And i think to myself, what a wonderful world

Re: Minimum memory size for zSeries Linux

ber is the second one: this represents the _total_ amount of memory available to Linux. mit freundlichem Gruß / with kind regards Carsten Otte -- IBM Deutschland Entwicklung GmbH Linux kernel coder Phone: +49 7031 16 4076 Tie Line: (

Re: Device to device copy within a running linux

(dasd=) and rerun zipl for the copy in order to make the copy runnable. In case you encounter problems, please CC: me directly in your response because I do not read this list frequently...! mit freundlichem Gruß / with kind regards Carsten Otte -- IBM Deutschland Entwicklung GmbH Linux kernel c

Re: Possible Bug in DASD Driver

but does not know (and cannot figure out without layering violation in Linux) whether devfs is mounted and used. Therefore it does always print both information: classic device name PLUS devno. I do not see any real issue here. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschlan

IUCV bugfix for linux-2.4.7.

M!!! >This has been fixed with this patch. > >I repeat: >DO NOT USE David's patch on guests with more than one CPU. >IT CAN CRASH YOUR VM. > >md5sum: >4a733eb1d48260f51eb0ed3bcb87cb72 iucv-2.4.7.patch > >-Fritz mit freundlichem Gruß / with kind regards Carste

Linux-2.4.7 CMS reserved minidisks fix

patches and fixes the problem: (See attached file: linux-2.4.7-extfix.diff) Note, that the problem is already fixed on 2.4.17, this is a retrofit of the corresponding patch! mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - device

Re: linux images hang or loop by turns

il by accident when reading the list. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - device driver team Phone: +49/07031/16-4076 IBM internal phone: *120-4076 -- We are Linux. Resistance indicates that you're missing

Re: Strange problem [Suse Linux on S390 VM]

a and some more messages - probably after the mark? Since we've got a long weekend in Germany this time, I'll try to decode the sense data on Tuesday, stay tuned. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - devi

Re: Multipath I/O on 390 Linux

mance than the workaround. The workaround however can help to boost access to a single volume with todays Linux distributions and without need to patch and rebuild anything because it helps to utilize different ESCON/FICON pathes in such a scenario. mit freundlichem Gruß / with kind regards Carsten Otte

Re: Multipath I/O on 390 Linux

) should fit best. This problem is already addressed in the current (experimental) 2.4.17 code. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - device driver team Phone: +49/07031/16-4076 IBM internal phone: *120-4076 -- We

Re: Antwort: Re: Antwort: AW: network problems during installation

Hi Tim! Try: insmod qeth.o "qeth_options=..." . Looks like the shell messes up your options string, the parantheses should protect it. Besides that, your options string looks fine AFAICS. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux f

Re: DASD API Version on 2.4.17

API) problem; instead of: if (MY_REQUIRED_API != KERNEL_API) problem; With the next currency drop we'll fix this one. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - device driver team Phone: +49/07031/16-4076 IBM internal

Re: dedicate DASD "not operational" under z/VM 4.1

y, VM folks are investigating this. mit freundlichem Gruß / with kind regards Carsten Otte IBM Deutschland Entwicklung GmbH Linux for eServer development - device driver team Phone: +49/07031/16-4076 IBM internal phone: *120-4076 -- We are Linux. Resistance indicates that you're missing the point!

<    1   2   3   4