Re: SLES15SP4 IPL failure: stage2 boot loader

[Christian Borntraeger]

Am 03.08.23 um 18:31 schrieb Clovis Pereira:

Looks like a problem with the host (KVM) not with the guest (SLES).


Right. There was at least one bug in QEMU where in (seldom) cases the boot 
loader failed:

https://gitlab.com/qemu-project/qemu/-/commit/5f97ba0c74ccace0a4014460de9751ff3c6f454a

And maybe the guest update to SP4 just triggered that corner case and the 
update to SP5 fixed it.
In any case qemu 5.2 from SLES15SP3 (or newer) should contain that fix so if 
you see that again,
updating the host should fix it.


Christian


Clovis Pereira
IT Mainframe Specialist - SW services
TLS - IBM Technology Lifecycle Services
55 11 99925-6242 (Mobile)
gclo...@br.ibm.com

IBM

De: Linux on 390 Port  em nome de Alan Haff 

Enviado: quinta-feira, 3 de agosto de 2023 12:31
Para: LINUX-390@VM.MARIST.EDU 
Assunto: [EXTERNAL] Re: SLES15SP4 IPL failure: stage2 boot loader


What QEMU do you have in the host?  (rpm -qa | grep qemu). Is that up2date?


qemu-s390-4.2.1-11.16.3.s390x

I know, I'm pretty far backlevel. Strange that it worked ok with SP3, though.

The good news is that I've been able to work around the problem for now by 
extracting the kernel and initrd from the guest disk and storing it on the host 
to IPL the guest with a direct kernel boot.

-Original Message-
From: Linux on 390 Port On Behalf Of Christian Borntraeger
Sent: Thursday, August 3, 2023 03:08
To: LINUX-390@VM.MARIST.EDU
Subject: [EXTERNAL] - Re: SLES15SP4 IPL failure: stage2 boot loader

Am 02.08.23 um 21:11 schrieb Alan Haff:

I have a SLES15 SP4 system running under KVM on a z13s that refuses to IPL after 
upgrading from SP3. The failure message is "! Cannot read stage2 boot loader 
!". I've IPLed into a rescue shell from the SP4 ISO, set up a chroot environment, 
and re-ran grub2-install with no obvious errors. Still no joy. Not sure what to try next.

Any ideas/suggestions/assistance would be greatly appreciated.


This looks like an error of the QEMU loader not being able to find the right 
things on the disk.
What QEMU do you have in the host?  (rpm -qa | grep qemu). Is that up2date?

--
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
lists...@vm.marist.edu with the message: INFO LINUX-390 or visit 
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: SLES15SP4 IPL failure: stage2 boot loader

[Christian Borntraeger]

Am 02.08.23 um 21:11 schrieb Alan Haff:

I have a SLES15 SP4 system running under KVM on a z13s that refuses to IPL after 
upgrading from SP3. The failure message is "! Cannot read stage2 boot loader 
!". I've IPLed into a rescue shell from the SP4 ISO, set up a chroot environment, 
and re-ran grub2-install with no obvious errors. Still no joy. Not sure what to try next.

Any ideas/suggestions/assistance would be greatly appreciated.


This looks like an error of the QEMU loader not being able to find the right 
things on the disk.
What QEMU do you have in the host?  (rpm -qa | grep qemu). Is that up2date?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Taking some time

[Christian Borntraeger]

Thank you Mark for all your work! I think we met first at a zExpo or SHARE
(around 2005?) and you have always been a person that did a lot of work
helping others. Very much appreciated!

All the best for your future life :-)

Christian

Am 15.02.23 um 16:55 schrieb Mark Post:

All,

I'm retiring from SUSE. Today is my last day on the job. I intend to
stay involved with Linux on the mainframe, including this mailing list,
but I'm not sure just how involved. I'll know better after I've had some
time to adjust to my new life.

When I left EDS in 2007, the environment at the company had become very
toxic, with layoffs every quarter of good people with decades of
experience. Joining Novell/SUSE at that time was the best possible thing
that could have happened to me professionally. It's been very satisfying
to contribute to the creation of SUSE's products. It has literally been
a dream job for me.

I'm thankful to the many people that have helped me over the years.
Without them, I might never have been able to make working with
mainframe Linux my career for 20+ years.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: zLinux 3270 console support

[Christian Borntraeger]

Am 14.11.22 um 22:11 schrieb Alan Altmark:

Sven asked about this on an internal IBM discussion channel and I mentioned 
that there are two issues that I can see in the Linux 3270 support:

1.  Linux is definitely doing it wrong on z/VM.  Linux infers the dimensions of 
the screen from a model number.  (Bzzt!)   If you have a standard-size 
screen for model 2, 3, 4, or 5, then the model number from DIAG 0x210 will be 
correct.  Otherwise it will reflect model 2.  I'm not sure why CP does that or 
why Linux is using the model number at all.   Even at the turn of the century 
it was irrelevant.

2. Linux doesn't detect a reconnect event, so it doesn't know to re-query the 
device and adjust any 'views' that have been created.



To be honest while someone could invest time in here, I think the proper 
solution for Linux would be to improve z/VMs ASCII console support (e.g. make 
it virtual instead of attaching the LPAR one and provide a console server).

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Openshift SNO on Z (KVM)

[Christian Borntraeger]

Am 17.10.22 um 05:24 schrieb Grzegorz Powiedziuk:

Hello,
has anyone tried installing openshift SNO (single node) on Z?
I was able to install it on x86 using the official procedure from redhat.
But going through the same process with KVM on Z doesn't work well so far.

I know it is not supported etc. This is just for testing/learning and fun.


Just reached out to our OCP team and in fact they have not touched SNO on the
IBM zSystems platform so far. Smallest OCP entity they have tested and therefore
support is 3-Node-Cluster.
Is there any usecase you follow except for fun and learning? A compelling case
will help to get some focus on SNO.




I am asking for any tips and clues so I don't go too deep into the rabbit
hole
thanks!
Greg

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Moving LUNs using z/VM

[Christian Borntraeger]

Am 06.04.22 um 17:48 schrieb Rick Troth:

Use 'dd' to copy block device to block device. (The name reminds me of DDR
every time I use it!)


Right. The syntax is a bit odd, but for most things
dd if= of= bs= count=
works.

For disk copy
dd if=/dev/disk/by-id/ of=/dev/disk/by-id/new

should be all that you need.
Then you can even mount,chroot and fixup things when necessary. Before that you 
should then also do blockdev --rereadpt on the target disk so that your LPAR 
understands the new partion table. 1







On Wed, Apr 6, 2022, 11:44 Martha McConaghy 
wrote:


Christian,
Looks like I have a couple of Linux LPARs that I'm going to have to move
to the new storage too.  (I was hoping to be able to retire them, but no
such luck.)  The trick with the EDEV isn't going to work for them (LUNs are
too big).

What would be a good tool to use in Linux to do physical copies of LUNs?
My colleagues usually only deal with copying filesystems, like rsync, which
would not do the trick for me.

Martha


Martha McConaghy

Marist:  System Architect/Technical Lead

SHARE Association:  Secretary

Marist College IT

Poughkeepsie, NY 12601


From: Linux on 390 Port  on behalf of Christian
Borntraeger 
Sent: Wednesday, April 6, 2022 1:56 AM
To: LINUX-390@VM.MARIST.EDU 
Subject: Re: Moving LUNs using z/VM

Am 05.04.22 um 23:19 schrieb Martha McConaghy:

I have a few RHEL servers that run on z/VM but boot off of a direct

attached SAN LUN (not an EDEV or ECKD).  They reside on an old DS8870 and I
need to move them to a new DS8910.  (No PPRC, GDPS, etc.)  Is there a way
that I could use VM to copy these LUNs to the new storage?   I was looking
at DDR, but wasn't sure if the FB-512 type would work for these.  They
aren't CMS format disks, obviously, so that isn't an option.  I'm trying to
avoid having to attach them to a Linux server to do the work, but will if
that is the only option.  Since these are boot volumes, I have to copy the
boot partition and boot record, not just the filesystems.  So, a physical
copy would be the best.

I think your last resort (using a Linux server) is actually  not a bad
idea as direct attached FCP is usually faster than EDEV.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390



--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Moving LUNs using z/VM

[Christian Borntraeger]

Am 05.04.22 um 23:19 schrieb Martha McConaghy:

I have a few RHEL servers that run on z/VM but boot off of a direct attached 
SAN LUN (not an EDEV or ECKD).  They reside on an old DS8870 and I need to move 
them to a new DS8910.  (No PPRC, GDPS, etc.)  Is there a way that I could use 
VM to copy these LUNs to the new storage?   I was looking at DDR, but wasn't 
sure if the FB-512 type would work for these.  They aren't CMS format disks, 
obviously, so that isn't an option.  I'm trying to avoid having to attach them 
to a Linux server to do the work, but will if that is the only option.  Since 
these are boot volumes, I have to copy the boot partition and boot record, not 
just the filesystems.  So, a physical copy would be the best.


I think your last resort (using a Linux server) is actually  not a bad idea as 
direct attached FCP is usually faster than EDEV.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: openJDK - openJ9

[Christian Borntraeger]

I asked our Java people and they answered the following:

I believe when you're using the api query that requests a specific version, for OpenJ9 builds you 
need to include the OpenJ9 version in the `release_name` field. If I change your url below from 
"11.0.13+8" to "11.0.13+8_openj9-0.29.0" it works.

https://api.adoptopenjdk.net/v3/binary/version/jdk-11.0.13%2B8_openj9-0.29.0/linux/s390x/jdk/openj9/normal/adoptopenjdk

Note that since July 2021, the OpenJDK+OpenJ9 builds at AdoptOpenJDK .net have 
been built by IBM using the same components and re-branded under the IBM Semeru 
Runtimes name.

For the Eclipse Adoptium API going forward, they will only "promote" certified 
builds, ie. IBM Semeru Runtimes Certified Edition. The Eclipse Adoptium Marketplace is 
still being worked on.

IBM has also done a "soft launch" on our own fork of their API which serves IBM 
Semeru Runtimes, both Certified Edition and Open Edition, which we host. We hope to have 
an announcement on that shortly. The code is the same so you can use ALMOST the same 
query and just swap out the base of the URL and the Vendor.
Instead of vendor adoptopenjdk use ibm for Open Edition and ibm_ce for 
Certified Edition.

Eg.
OE
https://ibm.com/semeru-runtimes/api/v3/binary/version/jdk-11.0.13%2B8_openj9-0.29.0/linux/s390x/jdk/openj9/normal/ibm
CE
https://ibm.com/semeru-runtimes/api/v3/binary/version/jdk-11.0.13%2B8_openj9-0.29.0/linux/s390x/jdk/openj9/normal/ibm_ce

Let me know if you have any questions.

Am 22.03.22 um 12:13 schrieb Neale Ferguson:

The adoptopenjdk.net site has been used to host builds of the JDK for both 
hotspot and openj9 builds (the latter being the IBM JVM which has many many 
optimizations for the Z environment). However, adoptopenjdk is moving to 
Eclipse’s adoptium site. However, according to the Eclipse site: “Eclipse 
Foundation projects are not permitted to distribute, market or promote JDK 
binaries unless they have passed a Java SE Technology Compatibility Kit 
licensed from Oracle, to which the OpenJ9 project does not currently have 
access.” I can’t pull the OpenJ9 JDK from that site. This is a PITA as there 
are multiple packages that pull JDKs as part of their build process. It means 
they can’t access the highly-optimized-for-z openJ9 JDK in this way (e.g. 
https://api.adoptium.net/v3/binary/version/).

It also seems that, even though the adoptopenjdk site still exists and I can 
download JDKs from there I cannot use the API to do things like: wget 
https://api.adoptopenjdk.net/v3/binary/version/jdk-11.0.13+8/linux/s390x/jdk/openj9/normal/adoptopenjdk
 whereas 
https://api.adoptopenjdk.net/v3/binary/version/jdk-11.0.13+8/linux/s390x/jdk/hotspot/normal/adoptopenjdk
 works fine. The “normal” appears to no longer be supported with s390x so 
“large” or similar should be used but substituting it in the former URL doesn’t 
work either.

Is there any expectation that openJ9 will be supported by Eclipse’s adoptium 
any time soon (i.e. before adoptopenjdk disappears)?

Neale


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: RHEL 6.10 installation under KVM failure

[Christian Borntraeger]

On 19.05.21 19:59, Alan Haff wrote:

I'm attempting to install RHEL 6.10 under KVM on a z13s and haven't been able 
to get the installer to IPL. When I try to IPL off of the ISO it fails with:

LOADPARM=[]
Using virtio-scsi.
SCSI CD-ROM detected.

! Cannot IPL this ISO image !

When I try to do a direct kernel IPL it fails with:

panic s390: core='0' psw-mask='0x000200018000' 
psw-addr='0x0002876e' reason='disabled-wait'

I've been able to IPL the RHEL 8 ISO and I've been able to do a direct kernel 
IPL of the SLES 11 SP4 installer. Any suggestions on why RHEL 6.10 won't IPL?


Two things here. The RHEL6 ISO is not bootable like the newer ones and RHEL6 
does
not have the necessary drivers to run under KVM (e.g. no s390x virtio drivers).
For KVM you need RHEL7 or later, SLES11 or later or Ubuntu 16.04 or later.

FWIW, RHEL 6.10 Maintenance Support II expired on Nov 30 2020 (basically it is
out of support) so do not expect any updates here.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: SYSASCII Console for Linux images

[Christian Borntraeger]

On 02.03.21 21:56, Dave Jones wrote:

Thinking about this a little more, I think what would be nice to have is
something along the lines of the OSA-ICC, which presents locally
attached 3270 device to the O/S, but is reachable via TCP/IP.

We could call it the OSA-ASC and it would present a locally attached
VT220 to either Linux running native in an LPAR or to Linux running as a
guest of z/VM, also reachable via TCP/IP.


This is basically the KVM model. We do provide virtual VT220 to the guest and
pass that along to local processes or network processes like virt-manager.

Not sure how hard that would be for z/VM as the model of operation is somewhat
tied to the 3270 console.


No need to have access to the HMC and it's Integrated ASCII console.


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: QEMU w/TCG: emulate missing CPU features?

[Christian Borntraeger]

On 01.03.21 18:52, Alan Haff wrote:
> Greetings,
>
> Will QEMU using TCG instead of KVM emulate CPU features that are missing on 
> the host processor? Specifically, will it emulate ESA/390 architecture on a 
> z14 or z15 with this config:
>
>   
> z900.3
>   
>
> for example?

I think TCG (like KVM) never implemented ESA/S390 but only zarchitecture.

>
> I've already confirmed that config doesn't provide features from newer CPU 
> models but we don't have a z14/z15 so I can't confirm that it will emulate 
> older features on newer CPUs.
>
> I know that performance will be worse (maybe much worse) with TCG instead of 
> KVM but I'm investigating this as a short-term solution for a specific use 
> case.

If you want to test some software if this is runnable on older hardware you can 
also use the CPU model with KVM.
While not perfect in every regard, the hardware provides a so call virtual 
architecture level that covers 2 additional generations.
This was introduced to provide live migration across different generations.
So when you for example use
zEC12
on a z14 then KVM will instruct the hardware to disallow instructions 
introduced with z13 and z14.
On a z15 you can go back to z13 for "perfect" fencing. If you go back to zEC12 
on a z15 then KVM
will instruct the hardware to fence everything from z14 and z15 and it will 
tell the guest that
only zEC12 facilities via stfle) are there. If code ignores stfle then it might 
be able to execution
some instructions, though.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Looking for advice setting up SMC-D between RHEL and z/OS

[Christian Borntraeger]

On 08.12.20 17:58, Eric Chevalier wrote:
> My apologies for the poor formatting of my original message. Let's try again:
> 
> [root@zlinux4 Temp]# hostname
> zlinux4.phx
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# date
> Tue Dec  8 06:37:35 PST 2020
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# lspci
> 00:00.0 Non-VGA unclassified device: IBM Internal Shared Memory (ISM) virtual 
> PCI device
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# ls -ld /sys/bus/pci/slots/*
> drwxr-xr-x 2 root root 0 Dec  7 11:23 /sys/bus/pci/slots/1018
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_rnics
>  FID  Power  PCI_ID    PCHID  Type   PPrt PNET_ID   
> Net-Dev
> -
>     1018  1  :00:00.0  07e1   ISM    n/a PNET1 n/a
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_rnics -e 1018
> Error: FID 1018 is already enabled
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# ip link
> 1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
> DEFAULT group default qlen 1000
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> 2: encf804:  mtu 1500 qdisc mq state UP mode 
> DEFAULT group default qlen 1000
>     link/ether 02:00:00:00:00:09 brd ff:ff:ff:ff:ff:ff
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_pnet
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_pnet -a PNET1 -I encf804
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_pnet
> PNET1 encf804 n/a 255
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_pnet -a PNET1 -D :00:00.0
> smc_pnet: Object exists
> [root@zlinux4 Temp]#
> [root@zlinux4 Temp]# smc_run sftp eric...@mvs60.phx
> eric...@mvs60.phx's password:
> Connected to eric...@mvs60.phx.
> sftp> quit
> [root@zlinux4 Temp]# smc_run -d sftp eric...@mvs60.phx
> libsmc-preload: map sock to AF_SMC
> eric...@mvs60.phx's password:
> Connected to eric...@mvs60.phx.
> sftp> cd Temp
> sftp> ls -l E*
> -rwxr-x---    ? 1050 0    429391872 Dec  1 13:14 EJES.V600.pax.Z
> sftp> get EJES.V600.pax.Z
> Fetching /u/ericadm/Temp/EJES.V600.pax.Z to EJES.V600.pax.Z
> /u/ericadm/Temp/EJES.V600.pax.Z 100%  410MB  29.9MB/s   00:13

With 30MB/sec you are far below what the network card could give you so this
is not the bottleneck. SMC cannot help you when your bottleneck is somewhere
else (e.g. disk I/O overloaded, subcapacity CPUs and unconfigured crypto 
resulting
in slow encryption). So I would suggest to watch the CPU load as well as the 
disk load to find out why things are slow.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: RHEL 8.3 install stuck

[Christian Borntraeger]

On 08.12.20 15:03, Rich Smrcina wrote:
> Friends,
> 
> We are trying to install RHEL 8.3 and the initial boot freezes at a specific 
> point.
> 
> Below is the parm file, the Q VSWITCH output for that machine, and the last 
> several messages of the console log.
> 
> I’m sure I’m missing something stupid, I just can’t see what it is.
> 
> ro ramdisk_size=4 cio_ignore=all,!condev  
>   
> ip=192.168.5.22::192.168.5.124:24:dsvmc.velocitysoftware.com:enccw0.0.0600:none
>  
> rd.znet=qeth,0.0.0600,0.0.0601,0.0.0602,layer2=0

Did you use that from a previous RHEL7? The device name has changed with RHEL 
enc600 instead
of enccw0.0.0600.
[..]
see here
> Ý7.399950¨ qeth 0.0.0600 enc600: renamed from eth0 
[...]

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Moving a Linux guest from z/VM to KVM?

[Christian Borntraeger]

On 01.12.20 12:07, Richard J Moore wrote:
> Is that really possible (KVM guest <-> z/VM guest)?
> As far as I am aware the KVM and zVM hypervisors share no metadata
> regarding the guest's set up. Nor do they share any of the protocols used
> in live (or dead) guest relocation.
>
> Am I misunderstanding the question?

I was not talking about live migration. I was more talking about:
site A: graceful shutdoen
size B: IPL

and this is possible as long as the dasd is available on both sides
and your network routing can tolerate this (or the guest can
tolerate to be in different networks depending on the IPL).

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Moving a Linux guest from z/VM to KVM?

[Christian Borntraeger]

On 30.11.20 21:10, Mark Post wrote:
> On 11/30/20 1:16 PM, Alan Haff wrote:
>> I have a number of Linux guests running under z/VM. I'd like to move some of 
>> them to a KVM host running in another LPAR. Naively I tried DDRing a guest's 
>> disks to new volumes and IPLing off of the new volumes. No luck, the virtio 
>> modules aren't available in the guest so it can't see any DASD.
>>
>> Has someone put together a recipe for moving a guest from z/VM to KVM? Or am 
>> I heading down a dead end?
>
> I'm not aware of one, but I wouldn't call it a dead end, either. As a
> first step, I would rebuild the initrd with whatever parameters are
> needed to include the virtio drivers, then try again. I have to think
> that /etc/fstab entries would need to be looked at as well.

Look at the virtualization cookbook 5 from the redbooks.
http://www.redbooks.ibm.com/abstracts/sg248463.html?Open

chapter 8.7 contains information how to configure a system as such that it can
be moved forth and back between LPAR and KVM. That should then also be possible
with z/VM. If done right that should allow you to move the guest forth and back
between KVM and z/VM.

Basic idea:
make sure the initrd has classic (DASD) drivers and the virtio-blk driver and
then use driver independent names for the partitions.

CC Viktor, do we still have our presentation flying around somewhere?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: performance problems db2 after moving from AIX

[Christian Borntraeger]

On 03.11.20 14:46, Grzegorz Powiedziuk wrote:
> Hi, I could use some ideas. We moved a huge db2 from old p7 aix to rhel7 on
> Z and we are having big performance issues.
> Same memory, CPU number is down from 12 to 10.  Although they had
> multithreading ON so they saw more "cpus" We have faster disks (moved to
> flash), faster FCP cards and faster network adapters.
> We are running on z114 and at this point that is practically the only VM
> running with IFLs on this box.
>
> It seems that when "jobs" run on their own, they finish faster than what
> they were getting on AIX.
> But problems start if there is more than we can chew. So either few jobs
> running at the same time or some reorgs running in the database.
>
> Load average goes to 150-200, cpus are at 100%  (kernel time can go to
> 20-30% ) but no iowaits.
> Plenty of memory available.
> At this point everything becomes extremely slow, people are starting having
> problems with connecting to db2 (annd sshing), basically it becomes a
> nightmare
>
> This db2 is massive (30+TB) and it is a multinode configuration (17 nodes
> running on the same host). We moved it like this 1:1 from that old AIX.
>
> DB2 is running on the ext4 filesystem (Actually a huge number of
> filesystems- each NODE is a separate logical volume). Separate for logs,
> data.
>
> If this continues like this, we will add 2 cpus but I have a feeling that
> it will not make much difference.
>
> I know that we end up with a massive number of processes and a massive
> number of file descriptors (lsof sice it shows also threads now, is
> practically useless - it would run for way too long - 10-30 minutes
> probably) .
>
> A snapshot from just now:
>
> top - 08:37:50 up 11 days, 12:04, 28 users,  load average: 188.29, 151.07,
> 133.54
> Tasks: 1843 total,  11 running, 1832 sleeping,   0 stopped,   0 zombie
> %Cpu0  : 76.3 us, 16.6 sy,  0.0 ni,  0.0 id,  0.0 wa,  1.0 hi,  3.2 si,
>  2.9 st
> %Cpu1  : 66.1 us, 31.3 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.3 si,
>  0.6 st
> %Cpu2  : 66.9 us, 31.2 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.3 hi,  1.3 si,
>  0.3 st
> %Cpu3  : 74.7 us, 23.4 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.3 hi,  1.3 si,
>  0.3 st
> %Cpu4  : 86.7 us, 10.7 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.3 si,
>  0.6 st
> %Cpu5  : 83.8 us, 13.6 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.6 si,
>  0.3 st
> %Cpu6  : 81.6 us, 15.2 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.9 si,
>  0.6 st
> %Cpu7  : 70.6 us, 26.2 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.9 si,
>  0.6 st
> %Cpu8  : 70.5 us, 26.6 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.6 hi,  1.6 si,
>  0.6 st
> %Cpu9  : 84.1 us, 13.6 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.3 hi,  1.3 si,
>  0.6 st
> KiB Mem : 15424256+total,  1069280 free, 18452168 used, 13472112+buff/cache
> KiB Swap: 52305904 total, 51231216 free,  1074688 used. 17399028 avail Mem

So you at least had some time where you paged out memory.
If you have sysstat installed it would be good to get some history data of
cpu and swap.

You can also run "vmstat 1 -w" to get an online you on the system load.
Can you also check (as root)
/sys/kernel/debug/diag_stat
2 times and see if you see excessive diagnose 9c rates.

>
> Where  can I look for potential relief? Everyone was hoping for a better
> performance not worse.I am hoping that there is something we can tweak to
> make this better.
> I will appreciate any ideas!

I agree this should have gotten faster, not slower.

If you have an IBM service contract (or any other vendor that provides support)
you could open a service ticket to get this analysed.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: We are expecting...

[Christian Borntraeger]

On 15.10.20 15:38, Frank M. Ramaekers wrote:
> ...a z15 and we currently are running on a zBC12.  So, I can I be sure that 
> my ClefOS will run just fine on my new machine (both have IFLs).
>
> cat /etc/redhat-release
> CentOS Linux release 7.8.2003 (AltArch)
>
> uname -a
> Linux host.domain.com 3.10.0-1127.18.2.el7.s390x #1 SMP Thu Jul 30 08:10:15 
> EDT 2020 s390x s390x s390x GNU/Linux

Not sure how much clefos differs from Rhel, but for RHEL z15 was tested with
RHEL 7.7: kernel-3.10.0-1058.el7, s390utils-1.23.0-45.el7; CCA lib: 6.0.13-10; 
EP11 lib: 2.1.0-8
see 
https://www.ibm.com/support/pages/ibm-z-linuxone-and-partner-certified-platform

So I would assume that this ClefOS kernel is newer and therefore good to go. 
(Neale to confirm)

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Overcommitting zLinux CPU

[Christian Borntraeger]

On 02.08.20 06:36, Alan Altmark wrote:
> 
> A physical core has a certain amount of “horsepower” in it.  It can, at top
> speed do X amount of work.
> 
> In SMT, you split the core in half, creating two execution contexts (CPUs)
> instead of just one.  The two CPUs share resources on the physical core,
> but the total horsepower doesn’t increase. In fact, it gets a little
> smaller in the sense that the core must now spend cycles managing the two
> CPUs (threads) on it.   Some workloads need more threads. Other workloads
> need faster CPUs. So you choose between SMT (threads) or non-SMT (speed).
> Knowing which is best means measuring workload response times.

To tell some more details: The sum of both SMT threads is usually larger
than one single thread. This is because a CPU does have many execution units
(floating point, fixed point, etc). Now the CPU can only execute things where
all dependencies are resolved.So several units are sitting idle, e.g. when there
was a wrong branch prediction until the pipeline has enough things in the out
of order window again. With SMT there are now 2 independent dependency tracking
streams that can make use of the execution units.

So as a rule of thumb: IF you have enough parallel threads and you are bound
by overall capacity, enabling SMT is usually a win for throughput. The z15 
technical guide says 25% on average over single thread. As an example that
could mean instead of 100% you get for example 60% + 65%.

What Alan tried to tell is that this of course DOES have an impact on latency.
When one single thread only gets lets say 65% the latency is larger.
So you balance latency vs throughput. And if you only have one thread on the
whole system, then this thread would be faster without SMT.

Now as latency might also depend on the questions "do I get ressources at all"
I also think that for highly virtualized systems with many guests and
overcommitment of CPUs, SMT is usually a win as z/VM or KVM have more threads to
distribute load. There is even some math for queueing systems that can show
reduced wait time for an idealized workload. 

There are cases where SMT is even worse, though. Some workloads really
go to the limit of execution units and if you have two of these workloads
splitting the overall number of lets say rename registers might actually hurt
performance so that the sum is smaller than just one single thread. The CPUs
got better over time and from z13->z14 and from 14 to z15 we identified several
of these cases and improved the CPUs. So on z14 and z15 SMT is a win most of 
the time.


[...]
> For these reasons, you generally do not want to have more virtual CPUs in a
> guest than you have logical CPUs to run them on.

Absolutely. Having more virtual than physical never makes sense apart from 
corner
cases like testing. But this statement is mostly for a single guest or a single
LPAR.

The sum of all virtual cpus can be higher as long as there is enough idle time,
e.g. not all cpus run 100% all the time.
For example with 4IFLs and SMT you can have 8 vcpus active at the same time. 
With
lots of idle systems that could also mean lets say 100 guests with 1vcpu each.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: SYSASCII console use for Linux

[Christian Borntraeger]

On 24.07.20 00:40, Davis, Larry (National VM Capability) wrote:
> I found this statement in the Linux-390 archives from Rick Troth and was 
> wondering if any one has used this interface or if there are more details on 
> using it
> 
> *Linux sees an ASCII "system console"*
> 
> This is where I had to do a little digging. They don't let me on the HMC
> so I have never gotten access to the SYSASCII interface. Sounds like a
> great idea. IBM realizes that some mainframe operating systems are
> better served with a byte-at-a-time interface. (Call it "ASCII", but the
> significant feature is not the character set. Character sets are so 1980.)

Yes, thats why KVM uses this type as the default console for its guests. Here
we have one virtual "ascii" console per guest that we then expose in a tty or 
pty.

> 
> z/VM can attach a SYSASCII to a Linux guest. Linux can then use the
> SYSASCII console interface and a non-mainframe Linux person would be
> perfectly comfortable.

Here it is different. You can attach the one "real" one to one guest. And 
the guest must see this console during boot if I remember correctly. 
Not sure if this will work with hotplug as the system console are initialized
very early. This will be especially hard as both consoles have the ttyS1 
_console_ name (the tty name of the ascii console is sclptty0). Maybe we 
should think about changing the console name as well to sclptty0.but
then this would break  existing setups that do use the console=ttyS1 parameter
in lpar and kvm.

 
> 
> We need a way to give our Linux admins real terminal access to an ASCII 
> terminal for Boot Messages and to fix issues that prevent Network access.
> 
> The Linux Admins do use an HMC access method for the AIX systems they support 
> and want to use the same thing for z/VM Linux servers
> 
> 
>   1.  Is this possible if the HMC has access to the z/VM LPAR
>   2.  Are there specific Steps to activate the SYSASCII console when attached 
> to Linux
>  *   Even if I attach SYSASCII to a Linux server it is not actually active
> 
> q sysascii
> 
> SYSASCII attached to GLTRHEL8 inactive
> 
>  *   How to I get SYSACSII activated
> 
> 
> 
> Larry Davis
> Senior z/VM systems Architect, Enterprise Services
> T +1.813.394.4240
> DXC Technology dxc.technology
> 
> 
> DXC Technology Company - Headquarters: 1775 Tysons Boulevard, Tysons, 
> Virginia 22102, USA.
> DXC Technology Company -- This message is transmitted to you by or on behalf 
> of DXC Technology Company or one of its affiliates. It is intended 
> exclusively for the addressee. The substance of this message, along with any 
> attachments, may contain proprietary, confidential or privileged information 
> or information that is otherwise legally exempt from disclosure. Any 
> unauthorized review, use, disclosure or distribution is prohibited. If you 
> are not the intended recipient of this message, you are not authorized to 
> read, print, retain, copy or disseminate any part of this message. If you 
> have received this message in error, please destroy and delete all copies and 
> notify the sender by return e-mail. Regardless of content, this e-mail shall 
> not operate to bind DXC Technology Company or any of its affiliates to any 
> order or other contract unless pursuant to explicit written agreement or 
> government initiative expressly permitting the use of e-mail for such 
> purpose. --.
> 
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
> 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: z/15 and SLES 11SP4

[Christian Borntraeger]

On 22.07.20 14:24, Christian Borntraeger wrote:
> On 22.07.20 13:07, Christian Borntraeger wrote:
>> On 21.07.20 23:40, Victor Echavarry wrote:
>>> Does anybody has an IBM z/15 running z/Linux SLES 11 SP4 
>>> kernel-default-3.0.101-108.87.1? we are in a middle of migration from SLES 
>>> 11 to 12 that supposed to finish at the end of year. Our company buy a z/15 
>>> and we want to know is the SLES 11 runs on the z/15 to finish the migration.
>>
>> The combination of distros and hardware that was tested by IBM and/or the 
>> distribution parter can be found here 
>> https://www.ibm.com/it-infrastructure/z/os/linux-tested-platforms.
>> As you can see SLES11 was not tested with z15. Reason was that SLES11 was 
>> already end of life (march 2019) when z15 went GA (see 
>> https://www.suse.com/lifecycle/) and only extended support contracts are 
>> available.
>> FWIW, if the planned software already supports SLES15, I would encourage you 
>> to switch to SLES15SP2 instead of SLES12. EOL for SLES12 is 2024 and EOL for 
>> SLES15 is 2028. But you need to double check that all dependencies are 
>> available for SLES15.
>
> Let me correct myself. We do not have a certification, but the combination 
> was tested inside IBM.
> So tested but not certified.
>
> See
> https://www.ibm.com/support/pages/node/6191619

Which also means that you should update your kernel to the version mentioned 
here or newer.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: z/15 and SLES 11SP4

[Christian Borntraeger]

On 22.07.20 13:07, Christian Borntraeger wrote:
> On 21.07.20 23:40, Victor Echavarry wrote:
>> Does anybody has an IBM z/15 running z/Linux SLES 11 SP4 
>> kernel-default-3.0.101-108.87.1? we are in a middle of migration from SLES 
>> 11 to 12 that supposed to finish at the end of year. Our company buy a z/15 
>> and we want to know is the SLES 11 runs on the z/15 to finish the migration.
>
> The combination of distros and hardware that was tested by IBM and/or the 
> distribution parter can be found here 
> https://www.ibm.com/it-infrastructure/z/os/linux-tested-platforms.
> As you can see SLES11 was not tested with z15. Reason was that SLES11 was 
> already end of life (march 2019) when z15 went GA (see 
> https://www.suse.com/lifecycle/) and only extended support contracts are 
> available.
> FWIW, if the planned software already supports SLES15, I would encourage you 
> to switch to SLES15SP2 instead of SLES12. EOL for SLES12 is 2024 and EOL for 
> SLES15 is 2028. But you need to double check that all dependencies are 
> available for SLES15.

Let me correct myself. We do not have a certification, but the combination was 
tested inside IBM.
So tested but not certified.

See
https://www.ibm.com/support/pages/node/6191619

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: z/15 and SLES 11SP4

[Christian Borntraeger]

On 21.07.20 23:40, Victor Echavarry wrote:
> Does anybody has an IBM z/15 running z/Linux SLES 11 SP4 
> kernel-default-3.0.101-108.87.1? we are in a middle of migration from SLES 11 
> to 12 that supposed to finish at the end of year. Our company buy a z/15 and 
> we want to know is the SLES 11 runs on the z/15 to finish the migration.

The combination of distros and hardware that was tested by IBM and/or the 
distribution parter can be found here 
https://www.ibm.com/it-infrastructure/z/os/linux-tested-platforms.
As you can see SLES11 was not tested with z15. Reason was that SLES11 was 
already end of life (march 2019) when z15 went GA (see 
https://www.suse.com/lifecycle/) and only extended support contracts are 
available.
FWIW, if the planned software already supports SLES15, I would encourage you to 
switch to SLES15SP2 instead of SLES12. EOL for SLES12 is 2024 and EOL for 
SLES15 is 2028. But you need to double check that all dependencies are 
available for SLES15.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM question.

[Christian Borntraeger]

On 30.06.20 22:19, Tom Huegel wrote:
> This has probably been discussed quite a bit but I wanted to ask for some
> real user feedback.
>
> I have a customer that is interested a POC of z/VSE zLINUX VTAPE.
> Unfortunately they do now have z/VM so the zLINUX would have to be in an
> LPAR. So the question is 'do I gain/lose anything using KVM as the
> hypervisor in the zLINUX LPAR?'.

As the zLINUX VTAPE application is just a Java applicaition that needs network
and disk access I do not see a reason why it could not run in a KVM guest. I 
guess
you have to ask yourself the question, will there be additional things that you
might want to run on zLinux to support VSE. I think you got plenty of ideas
in the recent mail threads. Then of course a hypervisor is going to help you.
One nice thing with KVM is that you can actually add that later. So having one
workload in an LPAR and then addding a KVM guest in the same LPAR to host 
another
one would just work. Not that this is something that I would suggest.
But the point is: no matter if you run Linux in LPAR or Linux under KVM. You
always start with a Linux in LPAR and can then move to KVM or not.

With some preplanning for the future you probably want to put the first workload
also in a KVM guest to have equivalent management for all workloads (e.g. use
CPU shares or capping).
Regarding the downside, of course every hypervisor will add a small overhead. It
starts with the memory footprint of an additional Linux kernel and it ends with
the cost of virtualizing I/O. Does this matter for the vtape use case? Probably
not.
As an alternative that depends on your requirements regarding isolation you 
could
put the workloads in containers. But this is only "nice" when the workload is
already containerized.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: z15 on-board compression

[Christian Borntraeger]

On 15.06.20 13:49, Christian Borntraeger wrote:

>> Wow more than 10x faster on dd - was not expecting that as I didn't think
>
> This looks more like an I/O difference.
>
>> it uses compression. But the compress with gzip -c, was only 25% faster on
>> the z15 while the decompress was about 4x.
>
> By default the hw is only used for compression level 1. So you can either use
> gzip -1
> or
> export DFLTCC_LEVEL_MASK=0x007e
> before using gzip.

Another thing. The exploitation is also not part of all distributions yet.
For example RHEL7 does not have it and for RHEL8 you need 8.2 and for Ubuntu
you need 19.10 or later.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: z15 on-board compression

[Christian Borntraeger]

On 10.06.20 01:36, Michael MacIsaac wrote:
> Hello list,
>
> I heard about the new DFLTCC instruction on the z15, aka on board
> compression.  I tried a quick experiment to see the difference from a z14.
> Disclaimer: I am not a performance expert.
>
> Here are three commands to create, compress and decompress a 1G file on a
> z14:
>
> # grep Type: /proc/sysinfo
> Type: 3906
>
> # time dd if=/dev/zero of=1G.file bs=1G count=1
> 1+0 records in
> 1+0 records out
> 1073741824 bytes (1.1 GB, 1.0 GiB) copied, 21.93 s, 49.0 MB/s
>
> real0m22.047s
> user0m0.001s
> sys 0m3.669s
>
> # time cat 1G.file | gzip -c > 1G.compressed.file
>
> real0m7.603s
> user0m5.362s
> sys 0m0.789s
>
> # time cat 1G.compressed.file | gzip -d > 1G.file
>
> real0m24.833s
> user0m4.103s
> sys 0m1.845s
>
> Here's the same commands on z15:
>
> # grep Type: /proc/sysinfo
> Type: 8561
>
> # time dd if=/dev/zero of=1G.file bs=1G count=1
> 1+0 records in
> 1+0 records out
> 1073741824 bytes (1.1 GB, 1.0 GiB) copied, 1.59126 s, 675 MB/s
>
> real0m1.621s
> user0m0.000s
> sys 0m1.216s
>
> # time cat 1G.file | gzip -c > 1G.compressed.file
>
> real0m5.722s
> user0m4.946s
> sys 0m0.510s
> # time cat 1G.compressed.file | gzip -d > 1G.file
>
> real0m6.150s
> user0m3.922s
> sys 0m1.290s
>
> Wow more than 10x faster on dd - was not expecting that as I didn't think

This looks more like an I/O difference.

> it uses compression. But the compress with gzip -c, was only 25% faster on
> the z15 while the decompress was about 4x.

By default the hw is only used for compression level 1. So you can either use
gzip -1
or
export DFLTCC_LEVEL_MASK=0x007e
before using gzip.


See also.

https://linux.mainframe.blog/zlib-acceleration/

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Redbook draft: The Virtualization Cookbook for IBM Z Volume 5: KVM

[Christian Borntraeger]

There is now a redbook draft for a KVM cookbook:
http://www.redbooks.ibm.com/Redbooks.nsf/RedpieceAbstracts/sg248463.html?Open

As this is still in draft state review feedback is still possible: 
redbo...@us.ibm.com

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: zLINUX end user provisioning

[Christian Borntraeger]

On 11.12.19 18:00, Alan Altmark wrote:
> On Wednesday, 12/11/2019 at 04:18 GMT, "Hamilton, Robert" 
>  wrote:
>> I'm interested in this too. I am following the read-only root 
> documentation,
>> which is starting to show its age a little (RedPaper redp4322 is good, 
> but
>> copyright 2008), but still has good stuff about gold instances, 
> maintenance and
>> cloning.
>> The other documents I have are the workbooks from some VM and Linux
>> installation classes I've taken, but a lot of the preparation stuff is 
> skipped
>> over.
>>
>> like, where and how to put an ISO. Nothing I've found in any cookbook 
> says how
>> it was done, or what tools I can use to manage or unpack an ISO. BFS? 
> OMVS? RPI?
>>
>> I'm just interested in the latest doc. I'm trying to install ClefOS and 
> manage
>> it as cleanly as possible.
> 
> Mike and I are in agreement about read-only root:  Don't do it. 
> Intellectually appealing, but impractical.  It's good for a ysprog to play 
> around with such things to see what's *possible* with the available tech, 
> but just because a thing CAN be done doesn't mean it SHOULD be done.

Agreed.
> 
> When we start to aggressively push process-changing Z tech into the Linux 
> admin world, it creates a rift.  The x86 Linux admins wash their hands of 
> Z, leaving it to a sysprog who isn't part of the Linux admin team to do 
> Linux administration.  And the sysprog is now left with another job and no 
> pay increase for doing it.  You don't get invited to the Linux admin 
> parties and you're not part of strategy discussions.  You get left behind 
> and sand gets kicked in your face.  You get exasperated eyerolls and the 
> occasional "OK, Boomer."
> 
> The real learning moment is when you simply ask your Linux admins, "How do 
> you install new Linux servers in your virtualized environments?"  Based on 
> that answer, you figure out how Z can fit into that model.  Just keep in 
> mind that better tech isn't always "best".  By all means, ask them if 
>  would be of value to them, but be prepared to accept 
> "No" as an answer, no matter how much inherent mainframe coolness is 
> there.

Right. It has to fit in the scheme of your company. 
I think a scheme that is used often in x86 is to use a cloud image and then
use cloud init to provision the system. You basically use a daily image
(e.g. https://cloud-images.ubuntu.com/bionic/current/ has daily builds
of latest greatest 18.04 with all updates). You then use a config file 
passed through as  virtual disk to provision the system. 
While this works natively with kvm, there are also documents that describe
cloud-init for z/VM. (I have not tested that with z/VM so I will just assume
that this will work as well) 

A kind of similar thing is used in coreos called ignition, so a base image 
that is customized with config seems to be common in the container/k8s
world as well.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Openshift Origin on KVM on LinuxONE

[Christian Borntraeger]

On 27.09.19 10:47, Johan Schelling wrote:
> I did some additional testing yesterday  using gdb and strace….
> 
> gbd didn’t return any useful information, but that might also be due to my 
> lack of gdb  experience.   Running strace resulted in the following:
> 
> ...
> clock_gettime(CLOCK_MONOTONIC, {207116, 975055264}) = 0
> clock_gettime(CLOCK_MONOTONIC, {207116, 975185579}) = 0
> mmap(0xc42020, 1048576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, 
> -1, 0) = 0xc42020
> mmap(0xc41ffe8000, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, 
> -1, 0) = 0xc41ffe8000
> clock_gettime(CLOCK_MONOTONIC, {207116, 975325872}) = 0
> clock_gettime(CLOCK_MONOTONIC, {207116, 975713954}) = 0
> mmap(0xc42030, 1048576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, 
> -1, 0) = 0xc42030
> mmap(0xc41ffe, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, 
> -1, 0) = 0xc41ffe
> clock_gettime(CLOCK_MONOTONIC, {207116, 977430825}) = 0
> getrandom(
> 
> Doesn’t matter whether I use docker 1.13 or docker-ce 18,  on the Clefos75 
> guest the daemon process hangs on the getrandom system call..
> 
> When I do the same on an Ubuntu 16.04 guest (where the daemon starts and runs 
> without any issues)  the strace shows the same getrandom call which executes 
> successfully:
> 
> ...
> futex(0xc420098948, FUTEX_WAKE, 1)  = 1
> futex(0xc42005e948, FUTEX_WAKE, 1)  = 1
> futex(0xc420098948, FUTEX_WAKE, 1)  = 1
> futex(0xc420098948, FUTEX_WAKE, 1)  = 1
> getrandom("h\"\277\352\376\262(\344", 8, 0) = 8
> clock_gettime(CLOCK_REALTIME, {1569355392, 476405161}) = 0
> clock_gettime(CLOCK_MONOTONIC, {2337380, 47849415}) = 0
> clock_gettime(CLOCK_REALTIME, {1569355392, 477070655}) = 0
> clock_gettime(CLOCK_MONOTONIC, {2337380, 48519415}) = 0
> clock_gettime(CLOCK_REALTIME, {1569355392, 477339454}) = 0
> clock_gettime(CLOCK_MONOTONIC, {2337380, 48781970}) = 0
> ...
> 
> Both the Clefos and Ubuntu guests are running in KVM.
> On the Clefos guest I have running in zVM  the getrandom call executes 
> successfully……
> 
> Reading up on the issue…. i found this:"When the entropy pool is empty, 
> reads from /dev/random will block until additional environmental noise is 
> gathered."
> Running  “cat /dev/random”  on the clefos  guest actually freezes….  Doing 
> the same on the Ubuntu guest does return data as does running the command on 
> the clefos guest in zVM .  So I feel that that’s where there problem is.
> Any other insights?

That makes sense. Can you maybe add a virtio-rng device to those guests
and install an rngd daemon in the guest that feeds the entropy from
/dev/hwrng back into the kernel entropy?

PS: some kernel versions had bugs where they did not get enough entropy from
disk and network activity. Maybe thats something to look at for Neale.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Openshift Origin on KVM on LinuxONE

[Christian Borntraeger]

On 24.09.19 11:13, Johan Schelling wrote:
> Goodmorning all,
> 
> I have been playing around for a while now with Openshift Origin  on our 
> LinuxONE system following the great “Getting_Started_with_OpenShift_v3.10”  
> guide….
> When using zVM as a hypervisor I can deploy an Openshift cluster with 
> multiple nodes with no problems at all (using zVM 6.4 and Clefos 7.5).  
> Everything works like a charm….
> 
> But when I try to deploy an Openshift cluster with one or multiple nodes 
> using KVM (using Ubuntu 16.04 KVM and Clefos 7.5) I run into problems 
> starting the docker daemon.
> Somehow the docker daemon freezes after which I have to cancel the ansible 
> playbook……   Starting the docker daemon by hand (systemctl start docker),  
> the same thing happens:


Can you say exactly what distro version run as host and what distro version 
runs as guest (the good and the bad variants)
This was not 100% clear.

> 
> ● docker.service - Docker Application Container Engine
>Loaded: loaded (/usr/lib/systemd/system/docker.service; enabled; vendor 
> preset: disabled)
>Active: activating (start) since di 2019-09-24 11:00:01 CEST; 8min ago
>  Docs: http://docs.docker.com
>  Main PID: 27563 (dockerd-current)
>CGroup: /system.slice/docker.service
>└─27563 /usr/bin/dockerd-current --add-runtime 
> docker-runc=/usr/libexec/docker/docker-runc-current 
> --default-runtime=docker-runc --exec-...
> 
> sep 24 11:00:01 lnxicu20 systemd[1]: Starting Docker Application Container 
> Engine...

any chance to attach gdb to that hanging process and doing a "thread apply all 
bt"?
> The service will remain in status “activating (start)”  until I kill the 
> service.  No messages (other than “starting Docker Application Container 
> Engine”) the logs.  Every docker command that you try to give (e.g. docker 
> version
> 
> Anyone seen this behaviour before or any ideas on what is going wrong?
> 
> I have a couple of other linux guests (both Ubuntu and SLES) running Docker 
> in both zVM and KVM environment without any problems….

So that means that you have other guests under KVM where docker starts up just 
fine. 

> 
> Some additional information:   yum list installed | grep docker
> 
> cockpit-docker.s390x176-4.el7.centos   @extras
> docker.s390x2:1.13.1-96.gitb2f74b2.el7.centos
> docker-client.s390x 2:1.13.1-96.gitb2f74b2.el7.centos
> docker-common.s390x 2:1.13.1-96.gitb2f74b2.el7.centos
> origin-docker-excluder.noarch   3.11.0-1.el7.git.0.62803d0 
> @centos-openshift-origin
> python-docker-py.noarch 1:1.10.6-9.el7_6   @extras
> python-docker-pycreds.noarch1:0.3.0-9.el7_6@extras

Or asked differently, is it just the clefos guest that does not work?
> 
> Same situation on linux guests running in zVM and KVM, but only on zVM the 
> docker service will start.
> 
> On the zVM guest the “docker version” command returns:
> 
> Client:
>  Version: 1.13.1
>  API version: 1.26
>  Package version: docker-1.13.1-96.gitb2f74b2.el7.centos.s390x
>  Go version:  go1.10.3
>  Git commit:  b2f74b2/1.13.1
>  Built:   Thu Jun 20 12:57:27 2019
>  OS/Arch: linux/s390x
> 
> Server:
>  Version: 1.13.1
>  API version: 1.26 (minimum version 1.12)
>  Package version: docker-1.13.1-96.gitb2f74b2.el7.centos.s390x
>  Go version:  go1.10.3
>  Git commit:  b2f74b2/1.13.1
>  Built:   Thu Jun 20 12:57:27 2019
>  OS/Arch: linux/s390x
>  Experimental:false
> 
> Running the same “docker version”  on the guest in the KVM environment 
> results in a freeze …..
> 
> On an Ubuntu guest running in KVM an “apt list —installed | grep docker” 
> returns:
> 
> docker/xenial,now 1.5-1 s390x [installed]
> docker.io/xenial-updates,xenial-security,now 18.09.7-0ubuntu1~16.04.5 s390x 
> [installed]
> 
> and the “docker version” command shows:
> 
> Client:
>  Version:   18.09.7
>  API version:   1.39
>  Go version:go1.10.4
>  Git commit:2d0083d
>  Built: Fri Aug 16 14:19:34 2019
>  OS/Arch:   linux/s390x
>  Experimental:  false
> 
> Server:
>  Engine:
>   Version:  18.09.7
>   API version:  1.39 (minimum version 1.12)
>   Go version:   go1.10.4
>   Git commit:   2d0083d
>   Built:Thu Aug 15 15:12:41 2019
>   OS/Arch:  linux/s390x
>   Experimental: false
> 
> The Ubuntu environment on KVM has been running perfectly OK for quite some 
> time now …..
> 
> Sorry for the long post, but this is driving me nuts…..
> 
> Regards
> Johan Schelling
> Infrastructure Solution Architect
> 
> 
> [cid:image001.png@01D008CB.2EEA8890]
> 
> 
> 
> ICU IT Services BV
> Transistorstraat 55b  I  1322 CK  ALMERE
> 
> M 06 – 21 245 992  I  E 
> johan.schell...@icu-it.nl
> T 088 – 5 234 123  I  www.icu-it.nl  I  KvK 32135776
> ICU 

Re: KVM and FCP disk without SAN switch

[Christian Borntraeger]

On 17.09.19 17:18, Johan Schelling wrote:
> I remember that when we started with KVM on LinuxONE around 3/4 years ago we 
> connected our V7000 directly (without a SAN switch) to the LinuxONE.  zVM 
> worked like a charm, but we had a lot of problems with (the then IBM 
> supplied) KVM…..  We had a lot of discussion with both the LinuxONE and the 
> storage guys as the documentation wasn’t quite clear on what was supported or 
> not. Problems we faced had to do with (hope I recall correctly) with WWPN’s 
> returned from the V7000 were  all mixed up… At the moment we added a SAN 
> switch all our problems were gone … ;-)
> Haven’t had any problems with KVM since …..   but haven’t checked the 
> situation without a SAN switch in the last years.

Yes, there was a firmware issue with the V7000s back then. But this has been 
fixed.
I still have a v7000 directly connected to my test z13 as an outcome of that 
bug back then.
But as I said, unless this is officially supported by all components you should 
better 
use a switch. 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Christian Borntraeger]

On 16.09.19 21:19, Jim Elliott wrote:
> For z/VM you must have a SAN switch to connect FCP attached disk. Is this
> also true for KVM (Ubuntu if it matters)?

As KVM borrows the FCP support from Linux the answer is the same as for Linux 
in LPAR.
In general the Linux code supports switched fabric and direct attachment. The 
code does
not support arbitrated loop.
What you also need is a support statement that tells about the storage server 
and the
Z system. For example a V7000 DOES work when connected to a z Box (there is no 
fencing)
but it is not a supported config (at least it was not when I check 3 years ago).
So in general you are on the safe-side with a switch. Without a switch requires 
to check
the certifications.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: NSS not possible in SLES 12

[Christian Borntraeger]

On 04.09.19 21:03, Rick Troth wrote:
> On 9/4/19 11:39 AM, Christian Borntraeger wrote:
>> On 04.09.19 16:41, Scott Rohling wrote:
>>> Let's start with who or what said it wasn't possible ?
>> [...]
>>>> Just to be sure, by "nss" I meant Named Saved System.
>> [...]
>>>>> what is the reason for nss not being possible with SLES from version 12?
>> [...]
>>
>> The Linux kernel now makes use of self-patching in several places and 
>> several core
>> features would no longer work without those.  To make NSS possible, the NSS 
>> would
>> need to have a copy-on-write semantics instead of being read-only. With 
>> global patching
>> we would copy almost everything over time making the feature not useful.
>>
>> So the feature was not only removed in SLES but will go away in other future 
>> distros
>> and it is no longer part of the upstream kernel.
> 
> 
> What's this? a little uptime funk? That's cool as long as it _doesn't
> break other things_.   
> 
> Seriously? You whacked NSS for live patching? Don't! (Too late.)  

Sorry I was not clear enough. 
This is NOT about the live patching in terms of updating your kernel. 

This is about changing the code of the Linux kernel during runtime for several
things like disabling trace points, applying CPU alternatives and choosing
security related instructions. Not having life patching for some of these things
would severely harm the overall performance as this would add additional 
branches
in too many places (e.g. every C function in the kernel) or even in places that
are not usable for a branch.
EVERYBODY (power,x86,arm,...)  now does live patching for these things that can
be dynamically enabled/disabled for a good reason and not doing so would prevent
us from using a big pile of these "smallish" features that will sum up over 
time.

 
> https://www.youtube.com/watch?v=SYRlTISvjww    
> 
> 
> Bad enough all the PUTTERING around in userland, even INIT, but now the
> kernel's borken too. Babies and bath-water both banished. Bummer!
> 
> 
> Hey, hey, hey, HAY ... Stop! ... wait a minute ... I'm a fan of advances
> (hallelujah!), but not at the cost of flexibility.
> 
> I believe y'all killed XIP too, right? That was brilliant. (NOT)

this is now called dax (direct access) and it still part of the dcssblk
device driver. I have not tested that recently though, so I can not say
that this still works but we have not removed that. 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: NSS not possible in SLES 12

[Christian Borntraeger]

On 04.09.19 16:41, Scott Rohling wrote:
> Let's start with who or what said it wasn't possible ?
[...]
>> Just to be sure, by "nss" I meant Named Saved System.
[...]
>>> what is the reason for nss not being possible with SLES from version 12?
[...]

The Linux kernel now makes use of self-patching in several places and several 
core
features would no longer work without those.  To make NSS possible, the NSS 
would
need to have a copy-on-write semantics instead of being read-only. With global 
patching
we would copy almost everything over time making the feature not useful.

So the feature was not only removed in SLES but will go away in other future 
distros
and it is no longer part of the upstream kernel.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Issues trying to install RHEL 8

[Christian Borntraeger]

On 31.07.19 04:52, Martha McConaghy wrote:
> Has anyone run into problems getting RHEL 8 to install on Z? I've run into a 
> weird problem.  Two different experienced Linux guys have looked at it, and 
> we are all stumped.
> 
> Background:
> 
> The install is being done in a 3G virtual machine on z/VM 6.4 on a LinuxOne 
> (z13 generation).  We have lots of different versions of Linux running on the 
> same system, including RHEL 7, CentOS, Ubuntu 18 and SLES 15.  So, I don't 
> think its a kernel/cpu type problem.
> 
> Problem:
> 
> When I run the REDHAT exec to bring up the install kernel, it starts booting 
> and things look like they will be OK.  Until it gets to the point where 
> "dracut initqueue" starts.  There is a message "7 urandom warnings(s) missed 
> due to ratelimiting".  It then pauses for a couple of minutes, followed by a 
> series of messages, "dracut-initqueue timeout - starting timeout scripts". 
> Eventually, it ends up in dracut in emergency mode.  The install process 
> never starts.  I've changed numerous parameters, even changed what type of 
> disk I'm using for the install.  It doesn't matter, its exactly the same 
> problem each time.

What exec and parmfile did you use? The network interfaces have different names 
now than REHL7
I think enc with  = devno. That can even be enc123 for devno 0123.

> 
> Really scratching my head on this one.
> 
> Martha
> 
> Ý"Ý0;32m  OK  "Ý0m+ Reached target Local File Systems.""
>  Starting Create Volatile Files and Directories...""
>  Starting Open-iSCSI...""
> Ý"Ý0;32m  OK  "Ý0m+ Started Open-iSCSI.""
> Ý"Ý0;32m  OK  "Ý0m+ Started Create Volatile Files and Directories.""
> Ý"Ý0;32m  OK  "Ý0m+ Reached target System Initialization.""
> Ý"Ý0;32m  OK  "Ý0m+ Reached target Basic System.""
>  Starting dracut initqueue hook...""
> Ý    6.306800+ dracut-initqueueÝ929+: RTNETLINK answers: File exists""
> Ý   84.168981+ random: crng init done
> Ý   84.168992+ random: 7 urandom warning(s) missed due to ratelimiting
> Ý  142.340863+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  142.931884+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  143.490965+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  144.080845+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> 
>  209.140407+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  209.700768+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  210.270398+ dracut-initqueueÝ929+: Warning: dracut-initqueue timeout - 
> starting timeout scripts""
> Ý  210.270531+ dracut-initqueueÝ929+: Warning: Could not boot.""
>  Starting Setup Virtual Console...""
> Ý"Ý0;32m  OK  "Ý0m+ Started Setup Virtual Console.""
>  Starting Dracut Emergency Shell...""
> Warning: /dev/root does not exist
> 
> Generating "/run/initramfs/rdsosreport.txt"
> 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Bad Superblock Expanding FS Redhat8

[Christian Borntraeger]

On 18.07.19 17:48, Frank Wolfe wrote:
> Good day all
>
> I am trying to expand my rootvg and did the following steps
>
> vmcp link '* 204 204 mr'
>
> cio_ignore -r 204
>
> chccwdev -e 204
>
> dasdfmt -b 4096 -d cdl -p /dev/disk/by-path/ccw-0.0.0204
>
> fdasd -a /dev/disk/by-path/ccw-0.0.0204
>
> pvcreate /dev/disk/by-path/ccw-0.0.0204-part1
>
> vgextend rhel_redhat8 /dev/dasde1
>
> pvdisplay /dev/dasde1  | grep "Total PE"
>
> lvextend --extents +586 rhel_redhat8/root
>
>
> Failing here:
>
> resize2fs /dev/rhel_redhat8/root resize2fs 1.44.3 (10-July-2018) resize2fs:
> Bad magic number in super-block while trying to open /dev/rhel_redhat8/root
> Couldn't find valid filesystem superblock.
>
> What step did I miss?

Is this maybe an xfs file system? Use xfs_growfs in that case.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: VoltDB on z?

[Christian Borntraeger]

On 07.01.2019 22:32, Jeffrey Barnard wrote:
> Neale,
> 
> CRC-32 is not the same as CRC-32c.
> 
> CRC32C uses a different polynomial (0x1EDC6F41, reversed 0x82F63B78) but
> otherwise the computation is the same. The results are different, of
> course. This is also known as the Castagnoli CRC32 and is found in newer
> Intel CPUs which can compute a full 32-bit CRC step in 3 cycles.

We do have both variants (crc32 and crc32c) as vector code in the kernel
> 
> The CKSM is CRC-32. We use the instruction to calculate IP, TCP and UDP
> checksums. I expect VCKSM to be CRC-32 also.

CKSM ís actually not CRC-32, it is a 4byte ones-complement addition that can
be used for IP, TCP and UDP. (TCP/IP does not use CRC!)

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Christian Borntraeger]

On 03.01.2019 19:32, Alan Haff wrote:
> I'm setting up a KVM-hosted guest in a SLES12 VM running under z/VM and I'm 
> not able to get networking going in the KVM guest.
> 
> Host: z/VM 6.4
> KVM host: SLES12SP4
> KVM guest: SLES12SP4
> 
> (The KVM host must run under z/VM; I don't have any control over that.)

For what its worth, this will have non-ideal performance, but it should work
anyway. Reason is that z/VM "emulates" the hardware virtualization (the
sie instruction) in that case. In fact, it actually tries to translate KVMs
sie control blocks into a shadow control block and runs the KVM guest as a
shadowed z/VM guest and fakes the result so that the kvm host "sees" what it 
would see on real hardware. It also has to shadow all page table actions of
the kvm host, so I strongly suggest to have enough memory (no paging) in the
KVM host to minimize the impact. Not sure about z/VM, but I would guess that
paging in z/VM also hurts as this also requires tear down of shadow page
tables - this is at least the case when running KVM under KVM.

We usually do not test this scenario, but I know that some
developers do have at least one development system with KVM under z/VM so
it gets some testing.

Not sure if you can talk about that, but I am interested in the use case.
I expect to have z/VM guests all over the place (in your case another z/VM 
guest)
or KVM only installations. Mixed mode only with different LPARs. Are there
interesting use cases that would suggest that we invest more time in making sure
that KVM runs well under z/VM?

[...]
> The macvtap0 device does get created when I start the KVM guest with 'virsh 
> start sles12':

as already said, can you use layer2?

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Christian Borntraeger]

On 08.01.2019 01:35, Robert J Brenneman wrote:
> KVM Host bridges require a L2 network interface with the 'bridge_role'
> attribute set on the OSA device supporting the bridge.
> ref: https://public.dhe.ibm.com/software/dw/linux390/docu/lhs4dd05.pdf
> chapter 14, section Layer 2 promiscuous mode  on p205.
> 
> But you're not using a L2 Vswitch - the 'q lan' response indicates it's a
> L3 : 'PERSISTENT RESTRICTED IP'
> if it had been a L2 network it would have said 'PERSISTENT RESTRICTED ETH'
> 
> 
> But wait - you're not trying to run a KVM Bridge , you're doing MACVTAP
> which attaches the whole eth interface to the KVM guest.

Well, not exactly. macvtap does not attach the whole eth, it registers one
mac address at that interface (via osa address table) and then passes along
the traffic for that mac address to the guest. Now this (as well as bridges
and openvswitch) always require a layer 2 interface.

If you really HAVE to use a layer3 interface only (e.g. a hipersocket that
connects to z/OS) you can do routing in the KVM host.

Examples of how to use that can be found in the paper
Exploiting HiperSockets in a KVM Environment Using IP Routing with Linux on Z - 
Results and Findings
https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/WP102746



> If you have the option to create a L2 LAN instead, try that first and see
> if it works.
> 
> If you can't create a L2 LAN to use for the MACVTAP interface, I am not
> sure whether the L3 LAN you have will work.  L3 LAN doesn't do ARPing at
> all so I'm not surprised wireshark is showing you no response to the
> outgoing ARP.  The virtual machine doesn't interact with the MACVTAP using
> the qeth driver internally, so you can't really put it in L3 mode to make
> it work that way.
> 
> I hope you have the option to create a L2 Guest LAN or vswitch for the KVM
> host's interface - the only other option I can think of would be to
> dedicate a triplet of OSA devices ( Read/Write/Data ) directly to the KVM
> Host's VM Guest.
> 
> On Thu, Jan 3, 2019 at 5:22 PM Alan Haff  wrote:
> 
>> I'm setting up a KVM-hosted guest in a SLES12 VM running under z/VM and
>> I'm not able to get networking going in the KVM guest.
>>
>> Host: z/VM 6.4
>> KVM host: SLES12SP4
>> KVM guest: SLES12SP4
>>
>> (The KVM host must run under z/VM; I don't have any control over that.)
>>
>> The KVM guest's network is defined as:
>>
>> 
>>   
>>   
>> 
>>
>> which, after 'virsh define sles12.xml', becomes:
>>
>> 
>>   
>>   
>>   
>>   
>>   
>>   
>> 
>>
>> The macvtap0 device does get created when I start the KVM guest with
>> 'virsh start sles12':
>>
>> 4: macvtap0@eth0:  mtu 1492
>> qdisc pfifo_fast state UP mode DEFAULT group default qlen 500
>> link/ether 52:54:00:93:f4:ce brd ff:ff:ff:ff:ff:ff
>>
>> Running WireShark on the KVM host when I try to ping the default gateway
>> from within the KVM guest, I can see ARP packets on the macvtap device (and
>> the transmit packet count increases) but I never see any ARP replies coming
>> back:
>>
>> macvtap0  Link encap:Ethernet  HWaddr 52:54:00:93:F4:CE
>>   UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1492  Metric:1
>>   RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>>   TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
>>   collisions:0 txqueuelen:500
>>   RX bytes:0 (0.0 b)  TX bytes:420 (420.0 b)
>>
>> I've confirmed that MACPROTECT is off:
>>
>> q lan
>> 00: LAN SYSTEM LANATTC  Type: QDIOConnected: 5Maxconn: INFINITE
>> 00:   PERSISTENT  RESTRICTEDIPAccounting: OFF
>> 00:   IPTimeout: 5 MAC Protection: OFF
>> 00:   Isolation Status: OFFVEPA Status: OFF
>> q vmlan
>> 00: VMLAN maintenance level:
>> 00:   Latest Service: VM65918
>> 00: VMLAN MAC address assignment:
>> 00:   System MAC Protection: OFF
>> 00:   MACADDR Prefix: 02DF02 USER Prefix: 02DF02
>> 00:   MACIDRANGE SYSTEM: 01-FF
>> 00:  USER:   00-00
>> 00: VMLAN default accounting status:
>> 00:   SYSTEM Accounting: OFF   USER Accounting: OFF
>> 00: VMLAN general activity:
>> 00:   PERSISTENT Limit: INFINITE   Current: 20
>> 00:   TRANSIENT  Limit: INFINITE   Current: 0
>> 00: Trace Pages: 8
>> 00: VMLAN Directory Network Authorization: ENABLED
>> 00: IVL Domain: None
>>
>>
>> Any ideas about what I could be missing?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Openstack on zlinux

[Christian Borntraeger]

On 10/09/2018 08:16 PM, Johan Schelling wrote:
> At ICU we run Openstack on Ubuntu on x86 and s390x (LinuxONE)…. build from 
> source using a couple of openstack-ansible playbooks. 
> Originaly the playbooks didn’t offer support for s390x but they were adapted 
> by my colleague Chris.  If you want to know more, you can contact him 
> directly on chris.beuk...@icu-it.nl 
> I’ll ask him if he can share some insights here…..

For Ubuntu you can also have a look at the cloud archive
https://wiki.ubuntu.com/OpenStack/CloudArchive.




> Johan Schelling
> Infrastructure Solution Architect
> 
> 
> 
> 
> 
> 
> ICU IT Services BV
> Transistorstraat 55b  I  1322 CK  ALMERE
>  
> M 06 – 21 245 992  I  E johan.schell...@icu-it.nl 
> 
> T 088 – 5 234 123  I  www.icu-it.nl   I  KvK 32135776
> ICU Proclaimer 
> 
>> Op 9 okt. 2018, om 19:01 heeft Jake Anderson  het 
>> volgende geschreven:
>>
>> Hi,
>>
>> Is there who has tried implementing openstack on zlinux .
>>
>> Could you point me to any manual which has brief steps to implement
>> openstack ?
>>
>> Jake
>>
>> --
>> For LINUX-390 subscribe / signoff / archive access instructions,
>> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or 
>> visit
>> http://www.marist.edu/htbin/wlvindex?LINUX-390
>> --
>> For more information on Linux on System z, visit
>> http://wiki.linuxvm.org/
> 
> 
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
> 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Alpine Linux 3.8 rc2 on z/VM and KVM

[Christian Borntraeger]

On 06/22/2018 05:18 PM, R P Herrold wrote:
> On Thu, 21 Jun 2018, Christian Borntraeger wrote:
> 
>> I can easily boot that iso image with kvm:
>>
>> $ qemu-system-s390x -enable-kvm -nographic \
>> -cdrom alpine-standard-3.8.0_rc8-s390x.iso
>> LOADPARM=[]
> 
> Hi, Christian,
> 
> What version of qemu-system-s390x are you running?

The KVM mode of qemu 2.12 build from source works fine.
For IBM z, the version in SLES12/15 Ubuntu 16.04 and 18.04
as well as RHEL 7-5 alt are fine.

> 
>   $ qemu-system-s390x -version
> 
> RHEL 7 derived seems to have a missing feature in glib-2 ; I
> have filed:
>   https://bugzilla.redhat.com/show_bug.cgi?id=1594304
> 
> [herrold@centos-7 ~]$ qemu-system-s390x -version

You seem to use the system _emulation_ of QEMU - and a very old
one. I have not tried this.

> 
> (process:17483): GLib-WARNING **: gmem.c:483: custom memory
> allocation vtable not supported
> QEMU emulator version 2.0.0, Copyright (c) 2003-2008 Fabrice
> Bellard
> [herrold@centos-7 ~]$

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Alpine Linux 3.8 rc2 on z/VM and KVM

[Christian Borntraeger]

On 06/21/2018 06:35 PM, Tuan M. Hoang wrote:
> On 06/21/2018 11:51 AM, Christian Borntraeger wrote:
>>
>>
>> On 06/21/2018 04:35 PM, Tuan M. Hoang wrote:
>>> Hi Christian,
>>>
>>> On 06/13/2018 03:32 AM, Christian Borntraeger wrote:
>>>> On 06/13/2018 05:58 AM, Tuan M. Hoang wrote:
>>>>> Hello,
>>>>>
>>>>> The Alpine Linux Project has released the the 2nd release candidate for
>>>>> its 3.8 stable series with support for z/VM and KVM.
>>>>>
>>>>> For the installation guide and more information, please visit :
>>>>> https://wiki.alpinelinux.org/wiki/S390x.
>>>>
>>>> Thanks for the good writeup.
>>>> Will there also be a bootable iso image to simplify the KVM installation 
>>>> even
>>>> more?
>>>
>>> rc8 includes a bootable iso image for KVM. Please try it out at:
>>> http://dl-cdn.alpinelinux.org/alpine/v3.8/releases/s390x/alpine-standard-3.8.0_rc8-s390x.iso
>>
>> I can easily boot that iso image with kvm:
>>
>> $ qemu-system-s390x -enable-kvm -nographic -cdrom 
>> alpine-standard-3.8.0_rc8-s390x.iso
>> LOADPARM=[]
>> Using virtio-blk.
>> ISO boot image size verified
>> Uncompressing Linux... Ok, booting the kernel.
>> [...]
>> Welcome to Alpine Linux 3.8
>> Kernel 4.14.51-0-vanilla on an s390x (/dev/console)
>> localhost login:
>>
>>
>> same happens with virt-install. So its an ready to use image but not an 
>> installer?
>>
> 
> Please login as root (no pwd) and motd will indicate running setup-alpine to 
> start the installer.
> 
> Alpine boot media (both netboot and ISO on all arch) are like a live cd 
> image, where you can choose to explore Alpine or run the installer.

Yes, works perfectly fine under KVM.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Alpine Linux 3.8 rc2 on z/VM and KVM

[Christian Borntraeger]

On 06/21/2018 04:35 PM, Tuan M. Hoang wrote:
> Hi Christian,
> 
> On 06/13/2018 03:32 AM, Christian Borntraeger wrote:
>> On 06/13/2018 05:58 AM, Tuan M. Hoang wrote:
>>> Hello,
>>>
>>> The Alpine Linux Project has released the the 2nd release candidate for
>>> its 3.8 stable series with support for z/VM and KVM.
>>>
>>> For the installation guide and more information, please visit :
>>> https://wiki.alpinelinux.org/wiki/S390x.
>>
>> Thanks for the good writeup.
>> Will there also be a bootable iso image to simplify the KVM installation even
>> more?
> 
> rc8 includes a bootable iso image for KVM. Please try it out at:
> http://dl-cdn.alpinelinux.org/alpine/v3.8/releases/s390x/alpine-standard-3.8.0_rc8-s390x.iso

I can easily boot that iso image with kvm:

$ qemu-system-s390x -enable-kvm -nographic -cdrom 
alpine-standard-3.8.0_rc8-s390x.iso 
LOADPARM=[]
Using virtio-blk.
ISO boot image size verified
Uncompressing Linux... Ok, booting the kernel.
[...]
Welcome to Alpine Linux 3.8
Kernel 4.14.51-0-vanilla on an s390x (/dev/console)
localhost login:


same happens with virt-install. So its an ready to use image but not an 
installer?


> 
>>>
>>> For running Alpine in Docker, please visit :
>>> https://wiki.alpinelinux.org/wiki/S390x/Docker.
>>>
>>> If you have any questions or comments, please reply to this thread or
>>> send an email to this list, or alpine-de...@lists.alpinelinux.org, or
>>> reach out at IRC channels #alpine-linux or #alpine-devel at Freenode, or
>>> use #alpinelinux on Twitter.
>>>
>  
> Tuan
> 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Alpine Linux 3.8 rc2 on z/VM and KVM

[Christian Borntraeger]

On 06/13/2018 05:58 AM, Tuan M. Hoang wrote:
> Hello,
> 
> The Alpine Linux Project has released the the 2nd release candidate for
> its 3.8 stable series with support for z/VM and KVM.
> 
> For the installation guide and more information, please visit :
> https://wiki.alpinelinux.org/wiki/S390x.

Thanks for the good writeup.
Will there also be a bootable iso image to simplify the KVM installation even
more?
> 
> For running Alpine in Docker, please visit :
> https://wiki.alpinelinux.org/wiki/S390x/Docker.
> 
> If you have any questions or comments, please reply to this thread or
> send an email to this list, or alpine-de...@lists.alpinelinux.org, or
> reach out at IRC channels #alpine-linux or #alpine-devel at Freenode, or
> use #alpinelinux on Twitter.
> 
> Many thanks,
> Tuan

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Linux "sleep" command not waking up under high CPU utilization

[Christian Borntraeger]

On 04/19/2018 04:56 PM, Dave Jones wrote:
> Thanks to all who responded to my query. At least we now know why the
> problem is occurring, and we will try some internal z/linux tuning.
>

Anything useful that can be shared with the list?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Linux "sleep" command not waking up under high CPU utilization

[Christian Borntraeger]

On 04/18/2018 03:36 PM, Dave Jones wrote:
> Hello, gang.
>
> I have a very simple bash script that runs a trivial data collection
> task, and then does a Linux "sleep im" to wait a minute before running
> the data collection task again. Under very high CPU loads (> 90%) I have
> noticed that the "sleep" command does not seem to wake up after one
> minute but instead wakes up 15 to 20 minutes later. This is on a Red Hat
> 6.9 guest running under z/VM 6.4 on a z12 box.
>
> I would like to buy a clue here if I could.

This should not happen, even under load. Is this really the sleep that does not
wakeup or is maybe the following stuff not getting access to the data? Have you
tried with "set -x" in the bash script to see what commands bash is executing?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: /proc/sysinfo on other architectures?

[Christian Borntraeger]

On 03/26/2018 03:38 PM, Michael MacIsaac wrote:
> Thanks to all who replied.
>
> So does anyone know how to query the hipervisor hierarchy, or if a Lintel
> is running on "bare metal"?

systemd-detect-virt

should work.

>
> -Mike
>
> On Mon, Mar 26, 2018 at 9:17 AM, Christian Borntraeger <
> borntrae...@de.ibm.com> wrote:
>
>> Yes, /proc/sysinfo is a method to expose parts of the STSI instruction to
>> unpriviledged code.
>> This is very z-specific.
>>
>> On 03/26/2018 02:38 PM, Michael MacIsaac wrote:
>>> Hello list,
>>>
>>> Is /proc/sysinfo a zLinux thing only?
>>>
>>> I got on a Lintel VMWare virtual machine and was surprised to not see
>> that
>>> file.  If it's not part of Lintel, how do they query their hipervisor
>>> hierarchy?
>>>
>>> Thanks.
>>>
>>> --
>>>  -Mike MacIsaac
>>>
>>> --
>>> For LINUX-390 subscribe / signoff / archive access instructions,
>>> send email to lists...@vm.marist.edu with the message: INFO LINUX-390
>> or visit
>>> http://www.marist.edu/htbin/wlvindex?LINUX-390
>>> --
>>> For more information on Linux on System z, visit
>>> http://wiki.linuxvm.org/
>>>
>>
>> --
>> For LINUX-390 subscribe / signoff / archive access instructions,
>> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
>> visit
>> http://www.marist.edu/htbin/wlvindex?LINUX-390
>> --
>> For more information on Linux on System z, visit
>> http://wiki.linuxvm.org/
>>
>
>
>
> --
>  -Mike MacIsaac
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: /proc/sysinfo on other architectures?

[Christian Borntraeger]

Yes, /proc/sysinfo is a method to expose parts of the STSI instruction to 
unpriviledged code.
This is very z-specific.

On 03/26/2018 02:38 PM, Michael MacIsaac wrote:
> Hello list,
>
> Is /proc/sysinfo a zLinux thing only?
>
> I got on a Lintel VMWare virtual machine and was surprised to not see that
> file.  If it's not part of Lintel, how do they query their hipervisor
> hierarchy?
>
> Thanks.
>
> --
>  -Mike MacIsaac
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Box confusion

[Christian Borntraeger]

First idea:
Is one guest part of a relocation domain with a z13 and the other one is not?
the cpu info is faked to the minimum machine in the relocation domain, (as this
is used by software to decide what to use) while sysinfo shows the real system.


On 01/24/2018 08:25 PM, Marcy Cortes wrote:
> I have a couple of SLES 12 SP3 servers on a new z14.   Both are running the 
> same levels of everything.
> Once of those correctly answers z14 to the "cputype" command.
> The other says z13
>
> myhostname:~ # cputype
> 2964 = z13 IBM z13
> myhostname:~ # cat /proc/cpuinfo
> vendor_id   : IBM/S390
> # processors: 1
> bogomips per cpu: 3111.00
> max thread id   : 0
> features: esan3 zarch stfle msa ldisp eimm dfp edat etf3eh highgprs 
> te vx sie
> cache0  : level=1 type=Data scope=Private size=128K line_size=256 
> associativity=8
> cache1  : level=1 type=Instruction scope=Private size=128K 
> line_size=256 associativity=8
> cache2  : level=2 type=Data scope=Private size=4096K line_size=256 
> associativity=8
> cache3  : level=2 type=Instruction scope=Private size=2048K 
> line_size=256 associativity=8
> cache4  : level=3 type=Unified scope=Shared size=131072K 
> line_size=256 associativity=32
> cache5  : level=4 type=Unified scope=Shared size=688128K 
> line_size=256 associativity=42
> processor 0: version = FF,  identification = 08B017,  machine = 2964
> myhostname:~ # cat /proc/sysinfo
> Manufacturer: IBM
> Type: 3906
> Model:400  M05
> 
>
> myhostname:~ # vmcp q cpuid
> CPUID = FF08B01739068000
>
> myhostname:~ # uname -r
> 4.4.103-94.6-default
>
> myhostname:~ # rpm -q s390-tools
> s390-tools-1.34.0-65.5.1.s390x
>
> Marcy
>
> This message may contain confidential and/or privileged information. If you 
> are not the addressee or authorized to receive this for the addressee, you 
> must not use, copy, disclose, or take any action based on this message or any 
> information herein. If you have received this message in error, please advise 
> the sender immediately by reply e-mail and delete this message. Thank you for 
> your cooperation.
>
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: How many Intel cores does an IFL emulate

[Christian Borntraeger]

As Jon Noltin already stated, z/OS does support SMT-2 for ZIIPs.
In addition to that, with z14 the system assist processors also run SMT-2 
enabled,
so we have IFLs, ZIIPs. SAPs that can enable SMT2.



On 11/09/2017 02:52 AM, Alan Altmark wrote:
> Thanks for advertising, Soup!  :-)
>
>  In fact, IFLs are the only engines on Z that have SMT-2, where each core in 
> the LPAR splits in two, appearing to the host as twice the number of CPUs.
>  Alan
>
>  Sent from my iPhone using IBM Verse
>
>  On Nov 8, 2017, 8:36:39 PM, soup...@gmail.com wrote:
>
>  From: soup...@gmail.com
>  To: LINUX-390@VM.MARIST.EDU
>  Cc:
>  Date: Nov 8, 2017, 8:36:39 PM
>  Subject: Re: [LINUX-390] How many Intel cores does an IFL emulate
>
>
>Come to think of it, I think an IFL is a single zSeries core.  z/VM can
>   fake it, though that's "overcommitting" the CPU resource.
>   Despite my NOT looking at the zSeries Principles of Operation I somehow
>   doubt the zSeries has hyperthreading...  as the z/VM OS would be dispatch
>   threads.  On the Intel CPUs, Linux is pretty smart in making a
>   hyperthreaded CPU "double" the number of dispatchable resources...  for
>   each "core" built into the die.
>   So I'm going to boast about my ignorance;  I don't think an IFL CP implies
>   multiple "cores" handling the zSeries instruction set.
>   -soup
>   On Wed, Nov 8, 2017 at 7:35 PM, Alan Altmark
>   wrote:
>   > On Wednesday, 11/08/2017 at 09:55 GMT, Philipp Kern
>   > wrote:
>   > > I thought unused CP capacity was literally costing OpEx whereas IFLs
>   > > would be CapEx? If it's both OpEx, then yes, agreed. (And I suppose
>   > > originally there might have been deals about giving some IFLs with
>   > > upgrades? ;-)
>   >
>   > Eh?  Processors are processors: CP, IFL, zIIP, CF, whatever.
>   >
>   > - Processor purchase: CapEx
>   > - Processor rental: OpEx
>   > - Processor recurring maintenance: OpEx
>   > - Software one-time-charge (z/VM): CapEx
>   > - OTC software recurring maintenance: OpEx
>   > - Monthly software charges (z/OS): OpEx
>   >
>   > This is why z/OS usage is so carefully monitored and controlled.  You can
>   > budget a large spike for a capital purchase, but uncontrolled monthly
>   > software charges create monthly budget crises.
>   >
>   > By adding capacity, you increase the software costs.
>   >
>   > If your software is priced per CPU (z/VM, Linux), then you will know that
>   > when you add a CPU, you also need to add money for licenses, and you know
>   > exactly what your maintenance bill will be.  Over-configured machines
>   > typically mean over-paying for your software and wasting money on
>   > maintenance charges on something you're not using.
>   >
>   > If your software is priced based on capacity rather than the number of
>   > CPUs, as z/OS is, then you typically look at subcapacity billing models
>   > that let you configure your hardware for spikes, but hold the rolling
>   > 4-hour average to a lower level.  There are mechanisms in z/OS that can be
>   > configured to thwart it's natural tendency (common to all OSes) to run as
>   > much as it can as fast as it using all the resources at its disposal.
>   >
>   > No rule applies 100% of the time (except this one), so there can be
>   > variations and combinations.
>   >
>   > Alan Altmark
>   >
>   > Senior Managing z/VM and Linux Consultant
>   > IBM Systems Lab Services
>   > IBM Z Delivery Practice
>   > ibm.com/systems/services/labservices
>   > office: 607.429.3323
>   > mobile; 607.321.7556
>   > alan_altm...@us.ibm.com
>   > IBM Endicott
>   >
>   > --
>   > For LINUX-390 subscribe / signoff / archive access instructions,
>   > send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
>   > visit
>   > 
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.marist.edu_htbin_wlvindex-3FLINUX-2D390=DwIBaQ=jf_iaSHvJObTbx-siA1ZOg=XX3LPhXj6Fv4hkzdpbonTd1gcy88ea-vqLQGEWWoD4M=eYU776321iiJg2XjRTylqK4_Sz9wjs0WKrKDOmwzhb8=VcV1mJ2gCYKMenPFgZ5fOGyenJ87wpav1R-3MTMwkRA=
>   > --
>   > For more information on Linux on System z, visit
>   > 
> https://urldefense.proofpoint.com/v2/url?u=http-3A__wiki.linuxvm.org_=DwIBaQ=jf_iaSHvJObTbx-siA1ZOg=XX3LPhXj6Fv4hkzdpbonTd1gcy88ea-vqLQGEWWoD4M=eYU776321iiJg2XjRTylqK4_Sz9wjs0WKrKDOmwzhb8=l6b7c2i0QG7nmc3lYnkNPEDNmbRK5ZGaZAhI7gGU5OY=
>   >
>   --
>   John R. Campbell Speaker to Machines  souperb at gmail dot
>   com
>   MacOS X proved it was easier to make Unix user-friendly than to fix Windows
>   "It doesn't matter how well-crafted a system is to eliminate errors;
>   Regardless
>   of any and all checks and balances in place, all systems will fail because,
>   somewhere, there is meat in the loop." - me
>   --
>   For LINUX-390 subscribe / signoff / archive access 

Re: RHEL 7.4 - Difficulties

[Christian Borntraeger]

On 11/06/2017 10:24 PM, Mark Post wrote:
 On 11/6/2017 at 03:16 PM,  wrote:
>> Hi,
>> Thank you for the response.  Umm... I'm not sure about this but I
>> circumvented with
>>
>>
>>
>>  echo 'GRUB_CMDLINE_LINUX=net.ifnames=0' >>/etc/default/grub
>>
>>
>> I would appreciate, you or anybody can bring more light into this.
>
> That's not a circumvention, that's the solution to your problem.  Unless Red 
> Hat decides to modify their kernels to use the old naming convention, this 
> will continue to be needed until you modify all your scripts to handle the 
> "predictable names."
>

An alternative (if you do not like to touch the kernel command line) seems to be
to run

ln -s /dev/null /etc/udev/rules.d/80-net-name-slot.rules

once.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vmcp cannot allocate memory

[Christian Borntraeger]

As I said, In the future vmcp will make use of the CMA allocator, which
should make sure that the allocation will succeed.  The advantage of that
allocator is that this memory is not kept free, but it can be used by movable
pages. So if vmcp needs that memory, CMA will be able to move that memory away
and free up the necessary contigous space.

FWIW, you can actually make sure that vmcp will succeed _today_ as long as you
keep the buffer size small enough. So as long as the buffer size is <= 2^3 pages
(32kbytes) things should always work.

So if you do vmcp inside a script or application you could split your queries 
into
smaller chunks and avoid the fragmentation issue, e.g.
instead of doing
Q -

do
q -07ff
q 0800-0fff
q 1000-17ff
q 1800-1fff
q 2000-2fff
...

I am aware of at least one case where we fixed that as such in an IBM product, 
but
I can not remember which one it was.


Christian

On 10/19/2017 08:27 AM, Pavelka, Tomas wrote:
> Would it be possible for the vmcp kernel driver to have an option to 
> pre-allocate some fixed size of contiguous memory and hold onto it? This 
> would prevent fragmentation which can hit you at any time. I remember using 
> vmcp from a Linux based installer which typically started by downloading 
> large files over scp. Somehow scp caused a lot of fragmentation which made 
> fragmentation errors very likely.
> If you have an application that runs a large number of vmcp commands, then 
> failed vmcp due to fragmentation means the end, because AFAIK there is no 
> sure way how to defragment, other than wait an unspecified amount of time.
>
> Tomas
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vmcp cannot allocate memory

[Christian Borntraeger]

Yes, this is memory fragmentation.
Heiko recently did some changes that should improve the situation.

This upstream change
https://git.kernel.org/pub/scm/linux/kernel/git/s390/linux.git/commit/?h=features=cd4386a931b6310b05559d2e28efda04d30ab593
will change the allocation from below 2GB to whereever is enough space left 
which should improve the situation for
larger guests.
We will try to integrate that into existing distro as a backport.

Long term we try to solve the issue in a more generic way:
Newer kernel contain a feature called contiguous memory allocator and the 
upstream commit
https://git.kernel.org/pub/scm/linux/kernel/git/s390/linux.git/commit/?h=features=3f4298427ad521fdc74fb991b17d84959513218a
does use this allocator for vmcp. The downside is, that this will be 
harder/impossible
to backport as older kernels miss the CMA infrastructure.

Christian


On 10/18/2017 07:50 PM, Neale Ferguson wrote:
> Fragmentation. Vmcp needs contiguous real storage. So if there is not a
> slab of 1MB in size hanging around then it will complain.
>
> On 10/18/17, 1:44 PM, "Linux on 390 Port on behalf of Michael MacIsaac"
>  wrote:
>
>> Hello list,
>>
>> We use vmcp extensively to issue CP commands.  More and more I'm seeing
>> this error:
>>
>> Error: Could not issue CP command: Cannot allocate memory
>>
>> I always use the --buffer=1M flag to be able to get the largest output.
>> I've
>> bumped up the VM size from 2 to 3 to 4GB, but still see it. We have some
>> VDISK swap:
>> # swapon -s
>> FilenameTypeSizeUsed
>> Priority
>> /dev/dasda1 partition   99156   0
>> 100
>> /dev/dasdc1 partition   1057796 0   1
>>
>> I reboot the system and it's gone.
>>
>> Has anyone seen this?  Seems like it could be considered a bug -
>> everything
>> else on the Linux system seems fine.  Or maybe I just need more swap?
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM's /proc/sysinfo

[Christian Borntraeger]

On 06/22/2017 02:37 PM, Michael MacIsaac wrote:
> Hello list,
>
> Could someone who has access to a KVM LPAR, please give me sample output
> from /proc/sysinfo for the hipervisor?
>
> This is what I get on a z/VM LPAR:

This is a z/VM guest not and LPAR, no?
>
> # grep ^VM00 /proc/sysinfo
> VM00 Name:ENGZS01
> VM00 Control Program: z/VM6.4.0
> VM00 Adjustment:  
> VM00 CPUs Total:  1
> VM00 CPUs Configured: 1
> VM00 CPUs Standby:0
> VM00 CPUs Reserved:   0
>
> ADVthanksANCE.

One of my sample guests has in /proc/sysinfo

VM00 Name:test137
VM00 Control Program: KVM/Linux
VM00 Adjustment:  1000
VM00 CPUs Total:  2
VM00 CPUs Configured: 2
VM00 CPUs Standby:0
VM00 CPUs Reserved:   0
VM00 Extended Name:   test137
VM00 UUID:4c3ae636-559d-4d90-b203-c8d3d150f0d0

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Debian CPU limitation?

[Christian Borntraeger]

On 03/22/2017 12:51 PM, Ray Mansell wrote:
> Thank you all for the information. It would appear we are stuck with 32
> CPUs for the moment and, since I'm retiring at the end of next week, I
> guess my replacement will have to decide how to proceed :-)
>
> Thanks again,
> Ray

As a last duty, can you maybe open a bug against Debian to bump the
max number of CPUs?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Debian CPU limitation?

[Christian Borntraeger]

On 03/21/2017 09:05 PM, Ray Mansell wrote:
> We've been running some beefy Debian servers, but of course one of them
> was not beefy enough, and we were asked if we could increase the number
> of CPUs from 32 to 48. Easy, I thought, until I rebooted the beefier
> server only to discover it had restricted itself to 32 CPUs, even though
> it knew 48 were present:
>
>cpu: 48 configured CPUs, 0 standby CPUs
>Brought up 32 CPUs
>
> The output from uname is:
>
>uname -a
>Linux linux02 3.16.0-4-s390x #1 SMP Debian 3.16.7-ckt9-3
>(2015-04-23) s390x GNU/Linux
>
> And it is running on a z/VM 6.3 host.
>
> I've searched through many documents and publications, but haven't been
> able to find anything that might help resolve this situation. Can anyone
> provide some insight, please?

Looks like the Debian kernel config for 3.16 contains
CONFIG-NR_CPUS=32.

This probably has made some sense in 31bit times (so that cpumasks in the
kernel could fit into a word size) but it looks like it was not increased
for 64bit to a higher value.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vector instructions

[Christian Borntraeger]

On 09/28/2016 05:22 PM, Richard Plangger wrote:
> Hi,
> 
> I'm working on pypy s390x port and I'm currently evaluating the vector
> instructions. I have one virtual machine access in marist.edu.
> 
> The vector optimization for PyPy do not show very good speedup for
> s390x. This is strange, because both PowerPC and x86 perform very well.

Do you have any examples of what you did? 
> 
> Is there some documentation about the setup (especially the vector
> facility) that one gets in a virtual machine?

You mean something that goes beyond what can be found in the principles of
operation for the z13? 

There are some changes that gave very good improvements in glibc (e.g.
the string operation ifuncs like
https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/s390/multiarch/strnlen-vx.S

or kernel (e.g. for crc32)
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/s390/crypto/crc32be-vx.S

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vdisk swap in sles 12

[Christian Borntraeger]

On 08/10/2016 07:48 PM, Marcy Cortes wrote:
...
>> Aug 10 11:55:32 zlnx169 swapon[3349]: swapon: /dev/dasdf1: read swap
>> header failed: Success Aug 10 11:55:32 zlnx169 kernel: Adding 92k

Another question. How did you prepare the swap disk, with mkswap in the guest 
or with
some CMS tool? Does a new mkswap on all failed disk + reboot help?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vdisk swap in sles 12

[Christian Borntraeger]

On 08/10/2016 07:11 PM, Marcy Cortes wrote:
> Working on VDISK in sles 12.
> We've always just run a /etc/init.d/boot.local on sles 11 but I decided to 
> try it the cookbook way.
>
> I have 4 disks.
> Only FF00 and FF02 show up, but swapon -a later will bring the other 2 online.
>
> This is in /etc/fstab
> /dev/disk/by-path/ccw-0.0.ff00-part1 swap swap pri=4 0 0
> /dev/disk/by-path/ccw-0.0.ff01-part1 swap swap pri=3 0 0
> /dev/disk/by-path/ccw-0.0.ff02-part1 swap swap pri=2 0 0
> /dev/disk/by-path/ccw-0.0.ff03-part1 swap swap pri=1 0 0
>
> Right after boot
> zlnx169:~ # swapon
> NAMETYPESIZE USED PRIO
> /dev/dasdd1 partition 976.6M   0B2
> /dev/dasda1 partition 244.1M   0B4
>
> They are OK, since swapon -a will bring them on
> zlnx169:~ # swapon -a
> zlnx169:~ # swapon
> NAMETYPESIZE USED PRIO
> /dev/dasdd1 partition 976.6M   0B2
> /dev/dasda1 partition 244.1M   0B4
> /dev/dasdc1 partition 488.3M   0B3
> /dev/dasdf1 partition   1.9G   0B1
>
>
> here are the relevant messages from journalctl
>
> Aug 10 11:55:32 zlnx169 systemd[1]: Found device 
> /dev/disk/by-path/ccw-0.0.ff02-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Activating swap 
> /dev/disk/by-path/ccw-0.0.ff02-part1...
> Aug 10 11:55:32 zlnx169 systemd[1]: Starting LVM2 PV scan on device 94:33...
> Aug 10 11:55:32 zlnx169 systemd[1]: Starting LVM2 PV scan on device 94:37...
> Aug 10 11:55:32 zlnx169 systemd[1]: Found device 
> /dev/disk/by-path/ccw-0.0.ff03-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Activating swap 
> /dev/disk/by-path/ccw-0.0.ff03-part1...
> Aug 10 11:55:32 zlnx169 systemd[1]: Starting LVM2 PV scan on device 94:25...
> Aug 10 11:55:32 zlnx169 swapon[3349]: swapon: /dev/dasdf1: read swap header 
> failed: Success
> Aug 10 11:55:32 zlnx169 kernel: Adding 92k swap on /dev/dasdd1.  
> Priority:2 extents:1 across:92k
> Aug 10 11:55:32 zlnx169 kernel: Adding 249992k swap on /dev/dasda1.  
> Priority:4 extents:1 across:249992k
> Aug 10 11:55:32 zlnx169 systemd[1]: Activated swap 
> /dev/disk/by-path/ccw-0.0.ff02-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: 
> dev-disk-by\x2dpath-ccw\x2d0.0.ff03\x2dpart1.swap swap process exited, 
> code=exited status=255
> Aug 10 11:55:32 zlnx169 systemd[1]: Failed to activate swap 
> /dev/disk/by-path/ccw-0.0.ff03-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Dependency failed for Swap.
> Aug 10 11:55:32 zlnx169 systemd[1]: Unit 
> dev-disk-by\x2dpath-ccw\x2d0.0.ff03\x2dpart1.swap entered failed state.
> Aug 10 11:55:32 zlnx169 systemd[1]: Found device 
> /dev/disk/by-path/ccw-0.0.ff00-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Activating swap 
> /dev/disk/by-path/ccw-0.0.ff00-part1...
> Aug 10 11:55:32 zlnx169 systemd-sysctl[3366]: Overwriting earlier assignment 
> of kernel/sem in file '/etc/sysctl.d/99-sysctl.conf'.
> Aug 10 11:55:32 zlnx169 systemd[1]: Activated swap 
> /dev/disk/by-path/ccw-0.0.ff00-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Starting LVM2 PV scan on device 94:17...
> Aug 10 11:55:32 zlnx169 systemd[1]: Starting LVM2 metadata daemon...
> Aug 10 11:55:32 zlnx169 systemd[1]: Started LVM2 metadata daemon.
> Aug 10 11:55:32 zlnx169 systemd-udevd[2006]: remove old symlink, 
> '/dev/disk/by-uuid/16795452-83d8-4b67-9a28-1e55947d9367' no longer belonging 
> to '/devices/cs
> Aug 10 11:55:32 zlnx169 systemd[1]: Found device 
> /dev/disk/by-path/ccw-0.0.ff01-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Activating swap 
> /dev/disk/by-path/ccw-0.0.ff01-part1...
> Aug 10 11:55:32 zlnx169 systemd[1]: 
> dev-disk-by\x2dpath-ccw\x2d0.0.ff01\x2dpart1.swap swap process exited, 
> code=exited status=255
> Aug 10 11:55:32 zlnx169 systemd[1]: Failed to activate swap 
> /dev/disk/by-path/ccw-0.0.ff01-part1.
> Aug 10 11:55:32 zlnx169 systemd[1]: Unit 
> dev-disk-by\x2dpath-ccw\x2d0.0.ff01\x2dpart1.swap entered failed state.
> Aug 10 11:55:32 zlnx169 swapon[3375]: swapon: 
> /dev/disk/by-path/ccw-0.0.ff01-part1: swapon failed: Invalid argument
>
> I may go back to the script.  But now I'm just too puzzled :)   How do i 
> figure out what a 255 is?

Is there a kernel message in "dmesg" ?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: "clever"(?) way to find files with duplicate contents.

[Christian Borntraeger]

On 07/06/2016 03:35 PM, John McKown wrote:
> I have a directory which has a number of files in it. I want to find out
> which files have identical content. Please, don't ask why (I'm an idiot?).
> Since these are text files, my first thought was to use diff. That is, list
> the files. For each file, do a diff against all the other files and note
> the result. I never came up with a decent algorithm to do this. Then I had
> a "vision". I remember that git stores file contents by basically creating
> a sha1sum, which it uses as a file name. Multiple files with the same
> sha1sum (which very likely to be unique based on the content) are only
> stored one. Now, since sha1sum is very unlikely to have a collision, how
> likely would sha512sum be to have a collision. So I did the following:
>
> for i in *;do x=$(sha512sum "$i" | cut -d ' ' -f 1);echo "$i"
>>> "${x}.sha512sum";done
>
> I then did:
>
> wc -l *.sha512sum | head -n -1 | awk '$1 != 1 {print $2;}'|while read i;do
> echo '===';cat $i;done
>
> which gave me a nice list of files with each group separated by ===.
>
> Is this reasonable? Is there a better way to do this?

Have you checked the "fdupes" tool?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Ubuntu 16.04 LTS Beta for IBM LinuxONE and z Systems

[Christian Borntraeger]

On 04/04/2016 08:01 PM, Neale Ferguson wrote:
> The z13 returns:
>
> 0xfb6bfffb 0xfcfff840 0x001c
>
> The bit string you are getting has bits off that are expected to be on.
> This missing bits are (words 2 & 3):
>
>  0800 0040

The kernel was fixed to not check for 73 (transactional execution) with
commit a69db2f6ad6915 ("s390/facilities: remove transactional-execution bits")
which is in 4.3 and should be part of 4.4 (ubuntu) as well.

>  0800 0040
  
This bit on the other hand (52) requires the Interlocked-Access Facility 2. 
This is
used to implement atomic bit operations when compiled for zec12 or later. So the
kernel refuses to boot if this facility is not available.
I cannot tell why zPDT does not claim support for this facility, though.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Does anybody know, what functionality SVC 29 does have on zLinux ? thx.

[Christian Borntraeger]

On 03/23/2016 12:30 PM, Leopold Strauss wrote:

SVC is the instruction that causes a supervisor call interrupt. This is
wired up in Linux to make a system call, in this case sys_pause

see
man 2 pause
or
http://linux.die.net/man/2/pause

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Status of VDISK after swap space usage

[Christian Borntraeger]

On 12/10/2015 05:58 AM, Alan Altmark wrote:
> On Wednesday, 12/09/2015 at 08:15 GMT, Barton Robinson
>  wrote:
>> My request to get this fixed was rejected by ibm several years ago. The
>> problem is that even though linux doesn't have anything on the vdisk,
>> z/vm still has to back it.  I asked for a diagnose as we do with real
>> storage - a way for Linux to tell z/vm the page no longer needs
>> backing.  So after the swapoff, can you detach the vdisk? That's the
>> only way with current technology to tell CP to free it up.
>
> The Linux community didn't like the memory state changes we asked for with
> CMM, so I don't see why they would assent to the change you describe.  A
> disk is just another address space that would need the same information.
> And there's no such thing as "I'm done" with swapping.

Newer Linux swap code does a discard (TRIM) on the full swap disk
if you do a swapon or on a bunch of pages if a cluster of that swap
disk becomes free. This depends on the parameters of swapon. This was
probably added to improve SSD handling. So if diag 0x250 would gain a
discard capability, something like that would become possible.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: zKVM on zpdt

[Christian Borntraeger]

Am 06.10.2015 um 18:32 schrieb Mark Post:
 On 10/6/2015 at 10:51 AM, Carsten Otte  wrote: 
>> Hi Tito,
>>
>> I am sorry that this time I don't have a workaround. The installer code
>> assumes diag 308 IPL functions to be available, however it is'nt on zPDT. I
>> believe this should be fixed in the installer. The Linux kernel propagates
>> the error to the script, and it fails to handle it.
> 
> So, is it safe to say that IBM didn't test this on zPDT and that IBM doesn't
> claim it is supported on zPDT?  Or that it positively is _not_ supported on 
> zPDT?  Having such a public statement might save others from some grief.
> 


zKVM is supported/tested on zEC12 and zBC12 and z13.
see
http://www-03.ibm.com/systems/z/solutions/virtualization/kvm/
and click on announcement letter.

When you read between the lines "Can co-exist with z/VM virtualization
environments, Linux on z Systems, z/OS®, z/VSE®, and z/TPF in different
LPARs." this also implies that zKVM is supposed to run in LPAR.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: How to force crash dump from hung Red Hat guest

[Christian Borntraeger]

Am 06.10.2015 um 17:41 schrieb Vitale, Joseph:
> Hello,
>
> Running  Red Hat 6.6  under zVM 6.3.  I had a guest hang and unable to issue 
> linux commands to force a kernel crash dump( echo c > /proc/sysrq-trigger ).
>
> Any suggestions how that can be accomplished?
>

If kdump is setup properly, then a PSW restart should do the trick.
On HMC/LPAR->recovery->PSW restart, on your z/VM the CP command it is #cp 
system restart

You can also setup the z/VM watchdog to do that.


A pretty good overview is here:
http://events.linuxfoundation.org/sites/events/files/eeus13_holzheu.pdf

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: zKVM installation problem

[Christian Borntraeger]

Am 05.10.2015 um 17:11 schrieb Ray Mansell:
> I'm trying to install zKVM in a virtual machine, but no matter what
> installation options I choose, I always get the following error:
>
> 2015-10-03 14:39:40,900 - controller.controller - INFO - InstallProgress
> screen
> 2015-10-03 14:39:40,901 - controller.controller - INFO - Formatting disks...
> 2015-10-03 14:39:40,901 - controller.controller - INFO - Installing KVM
> for IBM z into disk dasda...
> 2015-10-03 14:39:41,041 - model.installfunctions - INFO - Get repodata_file
> 2015-10-03 14:39:41,088 - model.installfunctions - CRITICAL - Failed
> installSystem
> 2015-10-03 14:39:41,089 - model.installfunctions - CRITICAL -
> EXCEPTION:
> 2015-10-03 14:39:41,089 - model.installfunctions - CRITICAL - local
> variable 's' referenced before assignment
> 2015-10-03 14:39:41,089 - model.installfunctions - CRITICAL -
> Stacktrace:Traceback (most recent call last):
>   File "/opt/ibm/kvmibm-installer/model/installfunctions.py", line 357,
> in installSystem
> installPackages(rootDir, callback)
>   File "/opt/ibm/kvmibm-installer/model/installfunctions.py", line 749,
> in installPackages
> repodata_file = getRepodataFile(repo, logger)
>   File "/opt/ibm/kvmibm-installer/model/installfunctions.py", line 548,
> in getRepodataFile
> d = re.split('([\d\w]+-primary.sqlite.bz2)', s)
> UnboundLocalError: local variable 's' referenced before assignment
>
> 2015-10-03 14:39:41,095 - controller.controller - CRITICAL - ZKVMError:
> [['KVMIBMIN70500', 'Error while installing packages.'], ('INSTALLER',
> 'INSTALLSYSTEM', 'INSTALL_MSG')]
>
> Help? Please?

Can you give Grzegorz suggestion of a preformat a try?
In case you have subscription & support, a PMR is certainly the right
thing to do to get this fixed properly.


Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: SLES12 + EDEV + bug

[Christian Borntraeger]

Am 01.10.2015 um 04:59 schrieb Grzegorz Powiedziuk:
> I think I might have found a small bug in latest update for SLES12  so this 
> is just a FYI for everyone who made the same mistake I did. 
> 
> If you use edevices, you know that the FBA driver  in linux automagically 
> (like Mark explained it to me few years ago ;) ) creates a device “1”  
> (dasda1 for example) on edevice. 
> You end up with for example dasda + dasda1  and you use dasda1 for OS 
> (including LVM) . No fdisks, no fdasds. 
> 
> I haven’t used edevices in a while so I forgot about that and my recent 
> SLES12 install I did on “dasda” (fba) , instead of  “dasda1”. I don’t know 
> how I did that but I did it and SLES install wizard didn’t complain. System 
> installed fine and it is was working ok.
> 
> pvscan on a build like this returned:
>  Volume Groups with the clustered attribute will be inaccessible.
>   PV /dev/dasda   VG root   lvm2 [19.53 GiB / 0free]
> 
> while in proper installation it looks like this:
> 
> PV /dev/dasda1   VG lnx15   lvm2 [19.53 GiB / 0free]
>   Total: 1 [19.53 GiB] / in use: 1 [19.53 GiB] / in no VG: 0 [0   ]
> 
> 
> Everything was working fine, until I did an update. Something has changed in 
> the way LVM recognizes physical  devices and it totally brakes the whole 
> system. It breaks all the parts where LVM is being called which includes the 
> little initrd which is loaded by  zipl during the first stage of boot. 
> 
> I did some debugging and I found that with latest update of SLES (I don’t 
> know why because the LVM seems to be in the same version) lvm doesn’t like 
> having metadata on “dasda” (fba) anymore. It likes it  only if its on dasda1 
> When pvscan scans the disk it ends up with:
>   /dev/dasda: Skipping: Partition table signature found 
> so it does not find the label and it fails to bring online the pv and volume 
> group. 
> 
> This edevice, if linked to an original SLES12 or any RHEL is working fine. 
> LVM finds a label on /dev/dasdb (fba) and I can mount it without a problem. 
> 
> I didn’t find anything different in lvm.conf which could cause this. 
> 
> To fix this (well it’s rather a dirty hack), I’ve downloaded a sourcecode of 
> LVM, found the instruction where it exits on message "Skipping: Partition 
> table signature found” , commented out that section, compiled, installed this 
> lvm, rebuilt initrds (both of them) and it worked. I got my system running 
> again. 
> 
> So the lesson is to make sure that SLES is being installed on dasda1 (fba) 
> not dasda   (only if you have edevices - with standard ECKDs  it’s probably 
> ok to do that) 
> 
> If you already have it on dasda  (FBA) - don’t update it without 
> preparations. 
> 
> Perhaps SLES shouldn’t allow to install a system this way at all?

Just guessing. Maybe this is a safety net to prevent the user from creating a
physical volume on an ECKD dasd without partition. For this case not going
onto a partion is indeed wrong. But doing so on an EDEV is certainly fine,
so maybe its just the check that is too unspecific.

Christian 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Documentation for Linux on z Systems and KVM - new

[Christian Borntraeger]

Am 24.09.2015 um 03:34 schrieb Grzegorz Powiedziuk:
> That’s very interesting. I wasn’t aware of this “stress” tool. So I’ve 
> downloaded it and run a couple tests with it.
> If I run basic —cpu 1 test (-n according to help is a dry run), the KVM 
> server spins the CPU 100% in user time. So no stealing at all. 
> 
> Could you run a couple of tests like this (I am providing my own results):
> 
> KVM server (2 CPU but it is one threaded task so doesn’t matter how many)

Does changing the KVM server to 1 CPU makes things better?
Having more than one CPU makes it harder for z/VM to virtualize
KVMs usage of SIE.




> # time for i in {1..500}; do  dd if=/dev/zero of=/dev/shm/test bs=1M count=10 
> ;echo interation $i done; done
> ….
> 10485760 bytes (10 MB) copied, 0.00273469 s, 3.8 GB/s
> interation 500 done
> 
> real  0m2.223s
> user  0m0.171s
> sys   0m2.002s 
> 
> During the test (I’ve changed 1..500 to 1..5000 to have more time to catch 
> top output) top was showing on average:
> %Cpu1  :  7.0 us, 83.7 sy,  0.0 ni,  7.0 id,  0.0 wa,  0.0 hi,  0.0 si,  2.3 
> st
> 
> 
> KVM virtual machine(1 CPU adding CPUs will not make difference in this case):
> # time for i in {1..500}; do  dd if=/dev/zero of=/dev/shm/test bs=1M count=10 
> ;echo interation $i done; done
> 10485760 bytes (10 MB) copied, 0.0524781 s, 200 MB/s
> interation 500 done
[]


> So it seems that when it is about only CPU the difference is very small or 
> none. I am getting similar results if I do complicated equations with big 
> numbers (virtual machines solves it almost in the same time as 1st level 
> host). 
> 
> But when memory is involved, everything slows down drastically. I wonder what 
> results you will get from the dd from /dev/zero to /dev/shm. 
> 
> And no, my system has plenty of memory, paging in z/VM is ZERO. Hardly 
> anything runs on this LPAR. 
> KVM host has plenty of real memory - 8G and the virtual machine is set to 4G. 
> It still has few Gigs left. No swapping, nothing else runs here. 

That certainly makes sense. Shared memory does use page protection for change
bit tracking. KVM does use page protection as well for dirty tracking of guests.
And if I remember correctly z/VM 6.3 also uses page table magic for 
change/reference
tracking. This is all fine. Unless: z/VM provides shadow page tables for the 
3rd 
level guest backing of a 2nd level hypervisor. It therefore needs to 
emulate/trap
most page table operations and page faults, which can be pretty expensive.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Documentation for Linux on z Systems and KVM - new

[Christian Borntraeger]

Am 24.09.2015 um 01:08 schrieb Mark Post:
 On 9/23/2015 at 05:57 PM, Grzegorz Powiedziuk  
 wrote:
>> As long as KVM can get close to z/vm performance then I
>> see a great potential in it.
>
> I don't think you're going to see that for quite a while.

I think we need to add some additional information.  All the announcement, GA,
support for SLES12 as guest talking from IBM and SUSE is not about the tech 
preview
in SLES12. It is about the IBM product
http://www-03.ibm.com/systems/z/solutions/virtualization/kvm/
together with the upcoming SLES12SP1 as a guest.

How much this differs from LPAR or z/VM or the KVM version in SLES12 certainly
depends on your workload. As always, pre-planning and testing is always a
good idea.

As of today the sweet spot of KVM on z is certainly a different one:
KVM provides  a different scheme of management and has different ways of
doing things. If you already have z/VM then you very likely want to
stay there as your operational model obviously works fine and its proven
and reliable in your shop.
If you already have KVM on x86 and no z/VM in the house, KVM on z might be
the  right tool to integrate the mainframe in your open environment because
it might integrate better into the operational model that is already
deployed.

> Those "8,000 virtual machines > on a z13" quotes I keep seeing from IBM are 
> all talking
> about z/VM, even though they never come right out and say it.  I would expect 
> far less
> capacity from KVM at this point in its development.

We have improved KVM since SLES12 was released. Some interesting 
reading/watching on
the progress of KVM and things we fixed in the past 2 years can be found here.

slides:
https://drive.google.com/file/d/0B6HTUUWSPdd-QlBta2ZEOWhQRlU/view?usp=sharing
presentation:
https://www.youtube.com/watch?v=cj-HLi1q6ZI

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Documentation for Linux on z Systems and KVM - new

[Christian Borntraeger]

Am 24.09.2015 um 00:42 schrieb Viktor Mihajlovski:
> On 23.09.2015 15:32, Grzegorz Powiedziuk wrote:
>> BTW, I was playing with KVM few days ago and it looks pretty awesome in 
>> terms of maintaining the environment and deploying new VMs but the 
>> performance for me was really bad.
>> And I mean extremely bad. I am not sure if it was because I made the KVM 
>> host (sles12) run as a virtual machine in z/VM or I was  doing something 
>> else wrong. I know that having kvm virtual machines in a 3rd level (under 
>> sles -> under z/VM) will impact performance but my case it was extremly bad. 
>> It was like running linux in hercules s390 in 2006 on old x86 desktop.
>>
>> The installation of linux in kvm virtual machine took 3-4 hours. Every 
>> operation that involves cpu and memory takes 3-10 time more time than on a 
>> KVM host itself.
>> Whenever something is happening in kvm virtual machine, the performance 
>> toolkit shows that KVM host is doing about 50% in supervisor mode and 50% in 
>> emulation mode which makes the t/v ratio for this machine about 2 which is 
>> pretty bad. I didn’t have time to do more investigation on this yet.
>> The KVM host (Server) sees about 50% cpu time as a “steal time”.
>>
> 
> Although KVM should generally be run in LPAR a slowdown in an order of 
> magnitude in z/VM seems a bit odd.
> I actually do run KVM under z/VM at my own risk (I am absolutely NOT 
> suggesting to do that). To recreate your problems I started a CPU burning 
> task in a KVM guest (stress -n 1) and see the following in the host:
> 
> Tasks: 162 total,   1 running, 161 sleeping,   0 stopped,   0 zombie
> %Cpu(s): 32,5 us,  0,7 sy,  0,0 ni, 63,8 id,  0,0 wa,  0,0 hi,  0,1 si,  2,9 
> st
> KiB Mem:812852 total,   801232 used,11620 free, 3152 buffers
> KiB Swap:  7212140 total,  1789268 used,  5422872 free,   611460 cached
> 
>   PID USER  PR  NIVIRTRESSHR S  %CPU %MEM TIME+ COMMAND
> 38763 qemu  20   0 2842280 353068 350952 S  94,8 43,4  58:46.16 
> qemu-system-s39
> 
> As you can see it's less than 3% steal time in the host with roughly 95% 
> problem state CPU utilization for the guest process (which indicates that SIE 
> isn't too bad even for a 3rd level guest).
> 
> The high steal time you observe could be a hint for either z/VM swapping on 
> behalf of the KVM host or KVM swapping itself. Could you observe a high page 
> rate?
> 

z/VM does call SIE on behalf of the guest hypervisor. So for CPU bound 
workload, which causes
almost no SIE exits things are fine. It is the sweet spot for 2nd level. As 
soon as the KVM
guest will have many exits (e.g. some I/O, memory fault-in, reschedule) or more 
than one CPU
in the KVM host things can get really slow as z/VM then has to interpret lots 
of things.
In addition z/VM 2nd level support was in no way optimized to speed up a KVM 
hypervisor, so
I assume that some of the optimizations for z/VM under z/VM have to fall back 
to the slow path.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: How to find a memory leak?

[Christian Borntraeger]

Am 10.07.2015 um 14:18 schrieb Bruce Hayden:
 The message sent to stderr is not documented in the device drivers book.
 It tells you about the response code of 2, but the description of that
 response code doesn't say anything about the error message to stderr or
 that the message tells you how long the output was.

I will have a look, if I can add something to the man page and the book.
At least the 2 message

Error: non-zero CP response for command 'command': #num
and
Error: output (num bytes) was truncated, try --buffer to increase size

should be documented, yes.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: How to find a memory leak?

[Christian Borntraeger]

Am 09.07.2015 um 19:16 schrieb Michael MacIsaac:
 I'm going to stop here for now.  I've learned a lot about Linux memory from
 this thread (but that's easy when you don't know much to begin with :)).

 I guess a question to the Linux developers in Germany would be:

 If vmcp is called with a buffer of 1M and the last slab in /proc/buddyinfo
 is 0, would it not be reasonable to nudge the kernel to free at least one
 slot up, assuming this can be done safely?

That does not help. The kernel frees up memory when needed (or when below a
watermark). So the AS-IS state does  not tell you anything. Now: newer kernels
(those that offer transparent hugepages) do have memory compaction which tries
to reorganize memory on pressure, so this case should be less of an issue.

Anothing thing: 1M is quite large (the larges contiguous memory that Linux wants
to allocate as slab). So try to not use that unless you need it. If you want to
know how much memory is needed, then vmcp gives you back the result of the 
diagnose:

# vmcp q dasd all
[...]
Error: output (74764 bytes) was truncated, try --buffer to increase size

# vmcp --buffer 75000 q dasd all
[...]
#

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: How to find a memory leak?

[Christian Borntraeger]

Am 10.07.2015 um 13:19 schrieb Christian Borntraeger:
 Anothing thing: 1M is quite large (the larges contiguous memory that Linux 
 wants
 to allocate as slab). So try to not use that unless you need it. If you want 
 to
 know how much memory is needed, then vmcp gives you back the result of the 
 diagnose:

 # vmcp q dasd all
 [...]
 Error: output (74764 bytes) was truncated, try --buffer to increase size

For those who want to trigger some action in that case:
The output from CP goes to stdout and the error message to stderr.
The return value of vmcp is 2 instead of 0 in that case. So something
if return == 2 then parse stderr and retry could work out.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Dynamically varying a CPU offline in ZVM

[Christian Borntraeger]

Am 24.03.2015 um 16:13 schrieb Bruce Hayden:
 You mark the CPU reserved (or really, give the number of reserved CPUs) in
 the LPAR profile.  Putting reserved CPUs and memory in the LPAR profile
 avoids IPLs!

On newer systems (does work on z196) you can dynamically add CPUs to the
LPAR profile and the running system, even without reserved CPUs - given that
you have enough CPUs on the system.

Christian

 On Tue, Mar 24, 2015 at 11:08 AM, Linker Harley - hlinke 
 harley.lin...@acxiom.com wrote:

 You should be able to vary if off from VM with

 VARY Online|OFFline PROCessor nn

 I don't remember exactly, but I think that there is a way to mark the CPU
 'reserved' so that it doesn't come online after an IPL.


 Harley



 --
 Bruce Hayden
 z/VM and Linux on z Systems ATS
 IBM, Endicott, NY

 --
 For LINUX-390 subscribe / signoff / archive access instructions,
 send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
 http://www.marist.edu/htbin/wlvindex?LINUX-390
 --
 For more information on Linux on System z, visit
 http://wiki.linuxvm.org/


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: sles 12

[Christian Borntraeger]

Am 30.10.2014 15:15, schrieb Levy, Alan:
 We currently have a z10 and z196. We have tried the sles12 install on both 
 systems and got the following error on both:

 LOADING SLES12 FILES INTO READER...
 DMSACP723I I (500) R/O
  NO FILES PURGED
 RDR FILE 0017 SENT FROM MASTER12 PUN WAS 0017 RECS 135K CPY  001 A NOHOLD 
 NOKEEP
 RDR FILE 0021 SENT FROM MASTER12 PUN WAS 0021 RECS 0002 CPY  001 A NOHOLD 
 NOKEEP
 RDR FILE 0025 SENT FROM MASTER12 PUN WAS 0025 RECS 339K CPY  001 A NOHOLD 
 NOKEEP
 DASD 0500 DETACHED
  NO FILES CHANGED
 The Linux kernel requires more recent processor hardware
 HCPGIR450W CP entered; disabled wait PSW 0002 8000  0BAD

 Any suggestions ?

This is a non-educated guess
Are both systems setup in a way to have migration between both? IIRC zVM might 
fake a z10 even on the z196 to allow for migration to z10.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: vmcp on RHEL 7?

[Christian Borntraeger]

On 10/01/2014 02:13 PM, Michael MacIsaac wrote:
 Hello list,

 Is vmcp is no longer a module on RHEL 7 (but it still works):

 # *modprobe vmcp*
 modprobe: FATAL: Module vmcp not found.
 # *vmcp q t*
 TIME IS 08:09:48 EDT WEDNESDAY 10/01/14

 Is this a change in the s390-tools, or just a RHEL 7 thing?

Its an kernel change. the vmcp kernel part is now always builtin.
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=f73a2b03c59b95a3ee8eebcc127350c77c950e87
This has changed with 2.6.35. (SLES11 has a change on top that reintroduces the 
modular build for SP1 to make this behave as the original GA version).

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: bogomips

[Christian Borntraeger]

On 21/02/12 21:18, William Carroll wrote:
 Isn't a BOGOMIP just a calculated loop value (ie how many times through
 the loop) used to do some internal timing.

It depends. On recent Linux versions (2007) the bogomips value is based on
the cpu capacity from store system information sysib 1.2.2.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: ext3...

[Christian Borntraeger]

On 15/12/11 19:09, Offer Baruch wrote:
 Hi,

 This is not a z/Linux or z/VM question but Let me ask you guys something
 about ext3... I guess that most of you are using it in production...
 How is it possible that i am using ext3 in my production systems and
 face stuff like:
 1. Corrupted FS during normal work that needs to be fixed with fsck or
 worse restore from a backup

Unless you do something wrong (typical candidates:
- mounting an ext3 at the same time on several systems
- file system on /dev/dasda instead of /dev/dasda1 on a CDL formatted disk
- logical volume on dev/dasda instead of /dev/dasda1 on a CDL formatted disk
- full disk backup of a running system from an external system
- misconfiguring the partition size after a resize
) or have broken hardware this should never happen.

 2. Resizing a FS requires me to fsck before I resize (as if the FS does not
 trust itself to be valid forcing me to umount the FS before a resize)
 3. Resizing a FS offline actually corrupts the FS

Cant tell about resizing.

 4. The fstab parameters, that states that it is normal to fsck your FS
 every boot or every several mounts...

You are the first z/OS admin that complains about too many sanity checks :-)
You can disable or change these values with tune2fs -c and -i.

 5. FS is busy although it is not mounted or in use by anyone...

This can also happen, if another file system is mounted on top

 6. fuser command will not always show the using processes

Might happen if there are short living processes or processes
that open/close all the time. (fuser does a snapshot and if these
processes are not accessing a file at that moment it doesnt work)

 7. open files can be removed without any warning from the rm command

This is actually a very nice unix feature. The directory entry is gone, but
the file does not go away (due to reference  counting on the inode) unless
the last user closes the file. And really its not a bug, its a feature -
e.g. programs can make sure that temp files got deleted when they crash.

 8. removing files from the FS will not free up space in the FS

Can be explained by 7. As soon as the last user goes away the space
will be freed.



Your 1. should really worry you. Again this should not happen. (and it
does not happen on my systems)

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: Assembler question

[Christian Borntraeger]

Am Mittwoch 07 Juli 2010, 10:12:12 schrieben Sie:
 My colleague is porting some assembler code to z Linux (gcc compiler  )
 and got some invalid op code assembler error for EPSW (extract psw)

gcc should handle epsw just fine.
And on SLES10 this works indeed:

# cat epsw.s
.globl main
main:
epsw %r2,%r3
svc 1  # exit system call. return value in r2
# gcc epsw.s
# ./a.out ; echo $?
1

By the way, if you are just interested in the condition code, then IPM + SRL
is faster.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: zLinux entropy

[Christian Borntraeger]

Am Samstag 01 Mai 2010 23:11:28 schrieb Greg Oldendick:
 On May 1, 2010, at 6:32 AM, Richard Troth wrote:

  But the crypto cards have their own API. I'n asking about the kernel's
  entropy function. Or is there a way to get the kernel to use the
  crypto?
 
 


 If you have the crypto card, you can load the prng.ko kernel module. It will 
 create a /dev/prandom device you can read from.

/dev/prandom is hardware supported pseudo random.
See Device Drivers, Features, and Commands page 297 (313 in acrobat)
http://public.dhe.ibm.com/software/dw/linux390/docu/lk33dd05.pdf

The real random numbers from the crypto cards is available via /dev/hw_random
See page 294 (310).

If your application needs to use /dev/random, I think there are tools or
daemons that feed entropy from hw_random into random.


Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: zLinux entropy

[Christian Borntraeger]

Am Montag 03 Mai 2010 16:36:20 schrieben Sie:
 I'm not seeing /dev/hw_random.

The z90crypt module must be loaded and your crypto card and the z90crypt
driver in your distribution must support the large random number feature.

The feature was introduced in 2008:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=2f7c8bd6dc6540aa3275c0ad9f657401985c00e9

Should be in SLES10 SP3 and RHEL5.3.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: SLES11 - shutdown seems to take forever

[Christian Borntraeger]

Am Mittwoch 24 Februar 2010 17:31:47 schrieben Sie:
 Hi Rich,
 It's the barebones install ( ala Cookbook ).  After this md: stopping all md 
 devices.,  you can count to about 11-12 Mississippi's ( I don't own a 
 stopwatch :-) before it actually comes down whereas under SLES9/SLES10 the 
 count is usually 1-3 Mississippi's.

There was a bug in the kernels udelay implementation that caused a long wait in 
the md code. As far as I know this fix
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=78d81f2f844b739b377817cfd279fb6067e191a7
will be part of SLES11 SP1.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: SIGILL problem on s390x

[Christian Borntraeger]

Am Freitag 20 November 2009 11:53:44 schrieb rui:
 Hi,

 The problem is, it happens at the start of the below mentioned function
 TDmain::Initialise (this=0x176c4c8) at crutil2.cpp:7571
 Before there was a call to srand(time(NULL)) there(the first line), i
 thought that was the problem, i have removed it now but it comes on the very
 first line of the function -- the call stack seems like corrupted, for i
 don't know what reason!

 cat /proc/cpuinfo
 vendor_id   : IBM/S390
 # processors: 1
 bogomips per cpu: 348.97
 processor 0: version = FF,  identification = 0ECA7E,  machine = 2096

2096 indicates a z9
[...]

 pc 0x7cc76e 0x7cc76e TDmain::Initialise()+70

The instruction before this address is certainly interesting.


 (gdb) print Initialise
 $1 = TDmain::Initialise()
 (gdb) x/i Initialise
 0x7cc728 _ZN7TDmain10InitialiseEv:   stm %r11,%r15,44(%r15)


Can you show the full disassembly of Initialize (use the disassemble command)?
A SIGILL can come for several reasons:
- opcode of newer machines (unlikely on a z9)
- specification exception
- data exception
and more.

In addition to the disassemble output, you can also do (as root)
 echo 1  /proc/sys/kernel/userprocess_debug

This will print some info in dmesg for exceptions in userspace.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: Converting EBCDIC with NEL to ISO8859-1

[Christian Borntraeger]

Am Freitag 20 November 2009 13:15:50 schrieb John McKown:
 I'm trying to figure out a better way to do this. Given: I have some
 EBCDIC text files on Linux. Instead of being delimited with 0x25
 (NewLine) instead of 0x15 (linefeed). This is how z/OS UNIX does text
 files and I cannot change it. When run through iconv the 0x25 is
 translated to 0x85, but I need 0x0a. I end up doing a tr '\205' '\n' in
 order to end up with 0x0a in the ASCII file. Note that when I do an
What codepage do you use for EBCDIC?
I think
iconv  -f IBM-1047 -t whatever Lnux code page, e.g. ASCII
should work for z/OS files. (untested)

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: SIGILL problem on s390x

[Christian Borntraeger]

 0x007cc758 _ZN7TDTmain10InitialiseEv+48:  lr  %r11,%r15
 0x007cc75a _ZN7TDTmain10InitialiseEv+50:  st  %r2,96(%r11)
 0x007cc75e _ZN7TDTmain10InitialiseEv+54:  lhi %r1,639
 0x007cc762 _ZN7TDTmain10InitialiseEv+58:  st  %r1,100(%r11)
 0x007cc766 _ZN7TDTmain10InitialiseEv+62:  l   %r1,100(%r11)
 0x007cc76a _ZN7TDTmain10InitialiseEv+66:  sfpc%r1,%r0
[...]
 User process fault: interruption code 0x40006
 CPU:0Not tainted
 Process crcat (pid: 12533, task: 00780878, ksp: 1140f4f0)
 User PSW : 0705e0008000 007cc76e (0x7cc76e)
 User GPRS:  0200027f 0176c4c8
 
807ccfda  
 74a8
00422d40 800ae330 40015be4
 7fffa540
02002000 0200807cc730 807ccfe4
 7fffa540
 User Code: 58 10 b0 60 a7 28 00 00 50 20 15 d0 58 10 b0 60 a7 38 00 00

Ok, so we have a specification exception for the sfpc instruction. Are there
some floating point instruction in this function?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: SIGILL problem on s390x

[Christian Borntraeger]

Am Freitag 20 November 2009 15:54:07 schrieb rui:
 Hi Christian,

 Yes, there is floating point precision related stuff, in the code on the
 very first line.

  unsigned int cm = 0x27f;
  _FPU_SETCW(cm);

Ok. That does the sfpc.

 This is done for all linux systems for achieving 64 bit floating point
 precision. So, do I need to ifdef it out for s390 or is there an alternative
 for it on s390x system?

So here is my understanding:
I strongly believe the mask for _FPU_SETCW is platform specific, the code
should use appropriate macros for floating point control. Your 0x27f is
based on the Intel numbers and will probably also have a different effect
on sparc, power, arm and mips (it might not fault, but could give you
strange results)

There are no precision bits for s390 in the floating point control. On s390
the precision depends on the instruction, and gcc should use double precion
instruction for variables declared as double.


 BTW, how did you match the spfc to user process fault, can you let me know
 about it, please?



 User process fault: interruption code 0x40006
 ^^
The 6 indicates an specification exception.

I got the sfpc  from the pc value in gdb + the disassembly:

pc 0x7cc76e 0x7cc76e TDmain::Initialise()+70
[...]
0x007cc762 _ZN7TDTmain10InitialiseEv+58:  st  %r1,100(%r11)
0x007cc766 _ZN7TDTmain10InitialiseEv+62:  l   %r1,100(%r11)
0x007cc76a _ZN7TDTmain10InitialiseEv+66:  sfpc%r1,%r0
0x007cc76e _ZN7TDTmain10InitialiseEv+70:  l   %r1,96(%r11)
0x007cc772 _ZN7TDTmain10InitialiseEv+74:  lhi %r2,0


On a fault, the pc usually points to the following instruction (or in rare cases
to the faulting instruction)

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: How do you . . . . setting up a user to force him to change his password?

[Christian Borntraeger]

Am Dienstag 13 Oktober 2009 17:31:01 schrieb CHAPLIN, JAMES (CTR):
 When setting up a new user, I want to be able to expire the user's
 temporary password, allowing the user to login with the temporary
 password, but force him/her to create his own password on the first
 login.

 I know how to set the INACTIVE value in the /etc/shadow file, but is
 there a command to modify the account without manually modifying the
 /etc/shadow create date (to expire) to do this?

I think the expire feature of passwd might be your friend:

# passwd -e userid

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: When will CMMA be removed from the kernel?

[Christian Borntraeger]

Am Mittwoch 23 September 2009 07:15:23 schrieb Mark Post:
  On 9/22/2009 at 11:35 PM, Leland Lucius lluc...@homerow.net wrote:
 
  Is it being removed entirely or will distros be providing it?

 First, let's distinguish between CMM-1 and CMM-2/CMMA.  It is the latter
  that is being dropped by IBM.  It will remain in the distribution versions
  in which it currently resides, as long as they are supported (I believe).
  Based on the comments from Shawn Wells, that means SLES only.

Yes, CMM-1 will stay.

You also have to distinguish CMM-2 full (SLES10) and CMM-2 lite.

We were not able to bring the full CMM-2 version upstream, but the lite version
is upstream. So the work in z/VM and hardware is not wasted.

CMM-2 lite also uses CMMA, so every recent kernel can use the CMMA facility.
CMM2-lite uses only two page states (used vs. unused) while the CMM2 in SLES10
uses 4 page states.

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: Which user env. variable tell me that it is in su - mode ?

[Christian Borntraeger]

Am Tuesday 03 March 2009 11:10:46 schrieb Marco Bosisio:
 Hi,
   do you know  wich user  environment variable I have to test  (executing
 a bash script)   to know  when  I am  in   su -   (switch user mode)   ?

Firstly, You should ask yourself why you need that information. If your script
behaves different in su - than a real root login, it is almost certainly
broken. Users expect that su - works like a real login.

You should check for any necessary capability instead.


Nevertheless,

In bash, something like
cat /proc/`cat /proc/$$/stat | cut -d   -f 4`/cmdline

gives you the command line of the bash's parent process.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: Which user env. variable tell me that it is in s u - mode ?

[Christian Borntraeger]

Am Tuesday 03 March 2009 13:58:59 schrieb Marco Bosisio:
 I customized the profile.local and it launch a script_xxx at login.  When 
 use  'su - '   I would like that the new user does not  execute the 
 script_xxx again.

Ah, ok. 
The thing is, profile is for login shells. And with - you _explicitely ask_
su to open a login shell. If you dont want that, run su without the -.

From man su:


DESCRIPTION
   Change the effective user id and group id to that of USER.

   -, -l, --login
  make the shell a login shell

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: Kernel bug in SLES 9

[Christian Borntraeger]

Hello Berry,

I have no advise on your specific problem, but a kernel bug, oops, warning or
panic is a very strong indication that this is a real code problem. I suggest
to open a service request/problem ticket.

hope this helps

Christian

 Message from /var/log/messages:

 kernel: kernel BUG at fs/aio.c:733!
 kernel: illegal operation: 0001 [#1]
 kernel: CPU:0Not tainted (2.6.5-7.315-s390x
 SLES9_SP4_BRANCH-200811261403180100)
 kernel: Process oracle (pid: 24794, task: 000144b81858, ksp:
 dd52b630)
 kernel: Krnl PSW : 07018000 001d2eaa
 (__aio_run_iocbs+0x1a2/0x39c)
 kernel: Krnl GPRS: 0008  001f
 00134d62
 kernel:001d2ea8  001d3b26
 
 kernel:001d39ca 2000 00013f7da300
 00013f7da300
 kernel:39bb8900 00378ac0 001d2ea8
 dd52bc60
 kernel: Krnl Code: 00 00 b9 04 00 2c b9 04 00 39 a7 49 00 00 c0 e5 ff ff
 fd d0
 kernel: Call Trace:
 kernel:  [001d39ca] io_submit_one+0x1b6/0x268
 kernel:  [001d3b26] sys_io_submit+0xaa/0x13c
 kernel:  [0011fc9c] sysc_noemu+0x10/0x16

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: question on top

[Christian Borntraeger]

Am Montag, 8. Dezember 2008 schrieb Barton Robinson:
 Yes, top lies.

This is no longer true with SLES10+ and RHEL5+. They use the stpt instruction
for accurate accounting.
If you still see wrong numbers with a recent distro, this would be a bug and
should be reported.

 So are you really saying a single threaded process is using more than one
 cpu?  Don't need much more proof than that.

Per default top accumulates all threads of a process. It is completely normal
to see values  100% if there are threads.
Use shift-H to see single thread values.

Christan

--
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: question on top

[Christian Borntraeger]

Barton,

 Sorry Christian, but with the latest and greatest, there are many cases where 
 Linux and
 TOP now seriously under report utilization (I think by factor of 5 in the 
 lab, and by 4 in

Do you have a short description of one test case? If there is a real problem,
we should fix it.

 a production server). Not sure we've bothered to report the details since 
 this problem
 would not impact our users.  So the data still can not be used for serious 
 performance

The last time we talked, your tool used the Linux data as one input value of
your calculations. So if the Linux data is really wrong, any fix would improve
the accuracy of your tool, no?

Christian

--
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 just can't git it

[Christian Borntraeger]

Am Montag, 6. Oktober 2008 schrieb Michael MacIsaac:
 # rpmbuild -bb git.spec
 error: Failed build dependencies:
 asciidoc is needed by git-1.5.6.5-9.1.s390x
 curl-devel is needed by git-1.5.6.5-9.1.s390x
 sgml-skel is needed by git-1.5.6.5-9.1.s390x
 xmlto is needed by git-1.5.6.5-9.1.s390x
 openssl-devel is needed by git-1.5.6.5-9.1.s390x

 So, using git is harder than I hoped it would be. Has anyone used git on
 s390x? Is there an easier way?  Thanks.

I think you must at least install openssl-devel and curl-devl from the DVD,
since git uses these packages to handle downloads.

Everything else is for the man pages and docu, so going to
http://git.or.cz/
and compiling the tar ball should give you something to use.

Christian

--
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 just can't git it

[Christian Borntraeger]

Am Montag, 6. Oktober 2008 schrieb Michael MacIsaac:
 -) Then using git failed, but at least it ran.
 # git clone http://git.gitorious.org/open-ovf/mainline.git
 Initialized empty Git repository in /usr/local/src/mainline/.git/
 fatal: http://git.gitorious.org/open-ovf/mainline.git/info/refs download
 error - transfer closed with 30 bytes remaining to read


Mike,

if possible, dont use http:// but use git:// instead, this gives a much better
performance and reliability. The git protocol understands the content and
transfers only the changes when doing a git fetch or git pull.

This is a tcp connection to port 9418.

--
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: swap to DCSS documentation/cookbook?

[Christian Borntraeger]

Am Freitag, 8. August 2008 schrieb Mark Perry:
 Pieter Harder wrote:
  Hi list,
 
  I am starting to think about the possibilities 2GB above the bar DCSS in
z/VM 5.4 will provide for replacing large Vdisk swap. But I can't find any
usable documentation on how to do swap to DCSS. The basics are documented on
dw and I can probably manage, but when anyone has been there and done that
can you please post some pointers and save us all the trouble of finding out
again. Performance information is quite welcome as well.
 
 Hi Pieter,
 DCSS is good for sharing code (or constant data) between Guests. Swap
 data is unique to only one Guest. What do you hope to gain over
 a/multiple VDISK - size? performance?

In theory swap over dcss can have lower overhead than vdisk, if z/VM has
enough memory to hold the dcss in main storage. For vdisks every I/O operatin
is an SIE intercept (either diag or channel program parsing). I/O to a fully
memory resident dcss I/O is memory access only and does not cause any sie
exit.

I have no idea about the performance, if z/VM has to do a lot of paging and
the dcss is not resident.

Christian

--
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: Hotplug agents

[Christian Borntraeger]

Am Mittwoch, 30. Juli 2008 schrieb Neale Ferguson:
 In SLES9 if I created a new device and device driver I could create an agent
 script and place it it /etc/hotplug such that devices could be automatically
 created. How does this work under SLES10?

udev does now all the processing which was done by hotplug. Look
at /etc/udev/rules.d/ for examples. For example you can use RUN+= to start
programs if some devices appear.

Christian

--
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: Hotplug agents

[Christian Borntraeger]

Am Mittwoch, 30. Juli 2008 schrieb Neale Ferguson:
 Tks.

 Should I just be able to add an entry to 50-udev-default.rules or create a
 new file with just this entry:


You can do both. But I prefer to create a new rule because then you can
package that rule with your device driver/program etc. For example
73-mdadm.rules is part of the mdadm rpm and not of udev. It also avoids
problems if any udev update from SuSE wants to replace 50-udev-default.rules
with a newer set.


 KERNEL=mydev*,  NAME=%k, MODE=666

 So that when the device driver gets up and creates entries in /sys for
 devices mydev0 etc. the rules will be triggered? (Assuming I've asked udev
 to re-read the rules files.)

Yes, if the device driver has done its registration with sysfs right (a dev
file comes up) it should work. In fact, if you dont care about the mode, the
default rule for udev should create the device node for mydev* without any
additional rule. (For example there is no rule for vmcp, it just gets the
defaults mode).

Christian

--
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: Hotplug agents

[Christian Borntraeger]

Am Mittwoch, 30. Juli 2008 schrieb Neale Ferguson:

 lrwxrwxrwx 1 root root0 Jul 30 13:43 bus - ../../../bus/iucv
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 local
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 msglimit
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 parmdata
 -rw-r- 1 root root 4096 Jul 30 14:15 path
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 priority
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 target_node
 -rw-r--r-- 1 root root 4096 Jul 30 14:15 target_user
 --w--- 1 root root 4096 Jul 30 14:15 uevent

Hmmm, there is no dev file in this folder, which is necessary for the basic
udev rule to work. The dev file is created when you device_register a
properly initialized struct device (the devt in this structure defines the
major and minor number and must be initialized to create the dev file). This
is the preferred method and it is probably only a small change in your
driver.

If you want to see what kind of events udev currently gets, open a 2nd shell
and start udevmonitor. Then load your driver in shell no. 1. udevmonitor
should list you all kernel events. In theory you could then hack a udev rule
that starts a shell script doing mknod, but I strongly suggest to wire up the
device_register (or device_create) stuff. The interface is not very pretty,
but it is preferred - even on SLES9.

Christian

--
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: Hotplug agents

[Christian Borntraeger]

Am Donnerstag, 31. Juli 2008 schrieb Neale Ferguson:
 I don't see a devt in the device structure - struct device in
 include/linux/device.h.

Yeah, right. I forgot. That was changed in 2006 or so. The devt was in struct
class_device in older kernels.So you have to register the class device as
well.

Christian

--
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: Using ESCON attached 3590 for TSM 5.5 / Linux

[Christian Borntraeger]

Am Mittwoch, 16. Juli 2008 schrieb Mike Hammock:
 They are running SLES 10 SP1 in an instance on a FLEX-ES system.  (Think of
 this as running in an LPAR on a normal system.  That is, no zVM.)
 They have an ESCON connected 3590 that they can use from the Linux system
 with no problem, but apparently there is a 'special' device driver supplied
 for TSM 5.5 and when they install it, the driver apparently does not 'see'
 the tape device.  They found a statement that TSM only supports Fiber
 attached tape devices, but we're not sure if Fiber in this context
 includes ESCON or not and I can't see why TSM should care as long as Linux
 is happy with the device.

 Does anyone else have any experience with this combination (TSM 5.5 on SLES
 10 with ESCON 3590) or possibly have any suggestions??

As far as I know, TSM for Linux only supports FCP-attached (scsi) tapes.
Channel attachement like FICON or ESCON are not supported.

Christian

--
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