that the userspace is 31bit, and thus a single
application cannot use more than 2G.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX
a chccwdev -d 0.0. even with an
udevadm settle --timeout=20 afterwards,
the file system gets corrupted, when I attach it to another guest.
Did you make sure to run sync before chccwdev?
(You could also use oflag=dsync but that would probably kill the copy
performance.)
Kind regards
Philipp
not recommended for
secure key generation.
Kind regards,
Philipp Kern
--
.''`. Philipp KernDebian Developer
: :' : http://philkern.de Stable Release Manager
`. `' xmpp:p...@0x539.de Wanna-Build Admin
`-finger pkern/k
regards,
Philipp Kern
[0] http://rhn.redhat.com/errata/RHBA-2011-0601.html
[1]
http://www.jur-linux.com/download/ftp.redhat.com/enterprise/6Server/en/os/SRPMS/s390utils-1.8.2-36.el6.src.rpm
[2] http://www.ibm.com/developerworks/linux/linux390/s390-tools-1.13.0.html
-march=z9-109 -mtune=z10.
Why is that, given that the architecture level set changed between z9 and z10,
not between z990 and z9?
Kind regards,
Philipp Kern
--
.''`. Philipp KernDebian Developer
: :' : http://philkern.de Stable Release Manager
On Tue, Aug 23, 2011 at 03:43:48PM -0600, Mark Post wrote:
On 8/23/2011 at 05:38 PM, Philipp Kern pk...@debian.org wrote:
Why is that, given that the architecture level set changed between z9 and
z10,
not between z990 and z9?
Which ALS are you referring to? From the hardware size
be.
Kind regards,
Philipp Kern
--
.''`. Philipp KernDebian Developer
: :' : http://philkern.de Stable Release Manager
`. `' xmpp:p...@0x539.de Wanna-Build Admin
`-finger pkern/k...@db.debian.org
On Thu, Sep 01, 2011 at 12:11:02PM +1200, Andrej wrote:
On 1 September 2011 10:39, Philipp Kern pk...@debian.org wrote:
That might be worth considering for Linux
You shouldn't do that there's the slightest possibility that you end up in
an
untrusted path (where somebody can drop
under z/VM 6.1 I also do not see 'edat' in /proc/cpuinfo, so hugepage
support is emulated in software.
And I can confirm that it's there on a z10 in LPAR mode. :)
Kind regards,
Philipp Kern
--
For LINUX-390 subscribe / signoff
one of their efforts to stop anyone else making GUIs,
smartphones, or whatever.
and then getting buyed by some evil corporation who uses the patents
against everyone else? (C.f. Sun and Oracle.) It's not that Sun was
particularly hostile to open source.
Kind regards
Philipp Kern
the one
you want to be in sync. Forcing it is kind of the point of the
[f][data]sync calls.
That said, the sent I/O requests might remain in the cache of the
storage and not actually be on any disk, but that's a different
matter I presume.
Kind regards
Philipp Kern
, but works just fine.
Probably not so well if your desktop IP address is NATed. Or am I missing
something?
CIFS doesn't have problems with NAT traversal. NFS does. CIFS is a
single TCP connection if you use IP-based URLs.
Kind regards
Philipp Kern
1503 ATTACHED TO ZLINUX1 1503 R/W NOASSIGN MULTIUSER
However, the lstape command shows no drives:
Did you set those devices online? Something like `chccwdev -e 0.0.1500'
might help.
Kind regards
Philipp Kern
--
.''`. Philipp KernDebian Developer
: :' : http
regards,
Philipp Kern
--
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
,
but single VMs?
Kind regards,
Philipp Kern
--
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
I found the advantage of PAV over single ECKD pretty severe with random read
workloads.
Kind regards,
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
compiler tool chain?
It's built from the normal gcc source, so yes. The Fortran compiler is written
in C, so as long as the backend supports the target architecture we should be
good to go. Anyway, R is shipped at least by Debian for z since ages.
Kind regards
Philipp Kern
of the corresponding IOCTL commands. You could try
installing a file system on the bare disk (no LVM) and see if that works.
At least on Debian you need to activate discard by setting issue_discards = 1
in /etc/lvm/lvm.conf's devices section.
Kind regards
Philipp Kern
regards
Philipp Kern
--
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
-ldap(d) in addition to
libpam-ldap(d).
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http
/post/2091092/a-year-with-mongodb
'nuff said.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http
? Because the OpenJDK runtime that's in
Debian is horribly slow on s390(x).
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO
you get normal virtio disks and NICs. SUSE has patches to
zipl to boot from virtio disks. On the Debian side we're still thinking
whether we should apply them or not, given that they won't be integrated
by IBM. (A silly upstream it is.)
Kind regards
Philipp Kern
the two, and access to up to date data as opposed to (what sounds
like) a copy of data being held in memory.
OTOH if you're burning CP cycles for them it might not be as good to get rid of
the caching layer. ;-)
Kind regards
Philipp Kern
On Tue, Oct 30, 2012 at 10:56:45PM +, OAKES Roger * ETS wrote:
And please don't let them talk about z/Linux in LPAR mode.
Why?
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions
.
They are dedicated Wintel people.
Debian will work on z9.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
to the reader. Punch files can only be 80 bytes, so the question
is 'How can the root= parm be continued?'.
Can't you simply continue on the next line? IIRC the lines 80 byte lines are
simply concatenated without spaces.
Kind regards
Philipp Kern
when it's available. For
the sake of Debian and others…
Thanks!
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http
probably not much as
the encryption will not use multiple threads.)
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390
Rob,
am Thu, Jan 31, 2013 at 03:14:09PM +0100 hast du folgendes geschrieben:
On 31 January 2013 14:38, Philipp Kern pk...@debian.org wrote:
Also you should be able to do between 100MB/s to 1GB/s on 10 GE, which is
My rule of thumb is that pumping 100 MB/s or so through the Linux
TCP/IP stack
-mount that ISO on the server and export
the result. You cannot loop-mount block devices over NFS.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
a heated topic, but still I hate to see unsourced comments
such as this.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390
. ;-)
Critique of the implementation can be very well warranted, but your initial
remark seemed like it aims at criticizing the New Way™ instead of the
implementation.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive
(4)).
You don't need to seed urandom. True, the quality of the output wouldn't
be great, but what's seen here should never happen and is clearly a bug.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive
. It is possible to issue a SysRq through
OS Messages anyhow?
Kind regards
Philipp Kern
[1] I cannot ask a vendor or open a PMR, in any case.
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists
Hi,
On Fri, May 03, 2013 at 10:38:49PM +0200, Philipp Kern wrote:
Does this ring a bell to anyone? Is there some known deadlock problem
that has been fixed in a later kernel? If not, does somebody know how
to debug that? I could try the dump process, but I'm not sure what I
could do
is the current version).
Kind regards
Philipp Kern
[1] http://www.debian.org/releases/stable/s390x/ch03s06.html.en#idp5389344
[2]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=installation-guide;dist=unstable
[3]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=installation-reports;dist=unstable
[4
-- www.livingcomputermuseum.org/
But I hadn't known about American Computer Robotics Museum --
www.compustory.com/
well, there's the Heinz Nixdorf MuseumsForum[1] in Paderborn, Germany, which
claims to be large. ;-)
Kind regards
Philipp Kern
[1] http://www.hnf.de/en/ueber-uns.html
tinkers with Debian on z, it will certainly
die. This may or may not be in your interest. :)
Kind regards and thanks
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists
). But then 3390-A are a DS8000-only
feature, I guess? I liked it being arbitrary-sized very much.
Of course there might be performance considerations, but not necessarily
when you then need to tie six 3390-3 together for space reasons.
Kind regards
Philipp Kern
.
Without having the code at hand I guess that s is only set in an "if"
but who knows. It should at least have been initialized with something.
At least you can just inspect the code given that it's Python.
Kind regards
Ph
.
Doing this, the failed logins went to zero. No more bots crawling around
and bruteforcing my VPS.
It should be enough to turn off PasswordAuthentication and
ChallengeResponseAuthentication to no in sshd_config and simply use
public key cryptography to login.
Kind regards
Philipp Kern
have very few signatures for Linux anyway.)
Yes, they say it's behavioral. I have yet to see a solution there that
works.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email
mething wrong in display vs. reply.
Kind regards
Philipp Kern
--
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.e
ryptoki comes from IBM ...
It might just be dependency bloat. opencryptoki is a more generic
PKCS#11 than just TPM as a backing storage. I think it supports
Crypto Express as tokens? And then it might just have inherited
the trousers dependency
nds on how you connect the storage on x86.
Kind regards and thanks
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
), which might impact your ability
to admin the system. (Not saying in a bad way, but different from what
people are used to on x86.) But is there any other significant
technology that'd save memory? Both are 64bit, although x86 has more
historical baggage around.
Kind regards and thanks
Philipp Kern
: If the application
allocates a lot of virtual memory but its working set is small and would
fit into RAM. There the size of the swap directly impacts if the memory
requests will be satisfied by the kernel or not. This behavior can be
tuned based on the information in [1].
Kind regards
Philipp
, and if so, could you share some details?
> If so, I could add it to the paper. Either on-list or off is fine.
The Debian/Ubuntu approach is preseeding the installation[1]. This is
not specific to System z either.
Kind regards
Philipp Kern
[1] https://w
Sometimes politicians think they need to
play with the time definition and there are updates to the file. If you don't
symlink you won't get them.
(This zone in question might be unlikely to change but people said that about
various timezones in the past and they changed, so...)
Kind regards
Phi
w s390-tools soonish.)
Kind regards and thanks
Philipp Kern
--
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
ow you get systemd-analyze.
Kind regards
Philipp Kern
--
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/wl
g
that, you can also introduce bastions that log all of their input/output
and that are the only way to contact the sensitive machines behind them
as a jump host.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / arc
ll have to decide how to proceed :-)
> As a last duty, can you maybe open a bug against Debian to bump the
> max number of CPUs?
This is now Debian bug #858731.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff /
comes down to the trade-off question if you're willing to
pay for the smart software and the smart brains to maintain it rather
than paying IBM to provide service for the mainframe.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe /
7bit ASCII.
But fix the mailing list, don't try to shard this out to the subscribers.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the me
course
need to get to the text representation by converting the
Content-Transfer-Encoding. In theory and in practice you also need to
normalize charsets across the mails. Which I suppose is skipped as well.)
Kind regards
Philip
well,
> you're going to have to minimize CPU overcommitment.
I idly wonder how actual utilization of IFLs looks like in the field.
It's clear to me that CPs run hot because otherwise you're wasting
moneys. I.e. if overcommitment is just a fact of life like it is in the
cloud.
Kind regards
Ph
Hi,
On 08.11.2017 16:42, Alan Altmark wrote:
> On Wednesday, 11/08/2017 at 03:17 GMT, Philipp Kern <p...@philkern.de>
> wrote:
>> I idly wonder how actual utilization of IFLs looks like in the field.
>> It's clear to me that CPs run hot because otherwise you'r
ll them on the same system. For binaries you can only have a
single variant installed. For libraries you can co-install multiple
architectures. The latter is actually how I cross-compile for s390x on
an amd64 system. And in theory you could even use qemu's user emulation
mode to run binaries compiled
ge table design and lookup/TLB logic in software.
You can maybe fake something on newer hardware but that'd be a large
feat for dubious benefits.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive acc
code
that cannot access more than 2GB of virtual memory. Is -m32 even a thing
on s390x?
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
space, and 64 bit binaries with
a 16 EiB address space, just like all competing
platforms provide.
x32 never took off either. And just because a platform provides it for
compatibility doesn't mean it makes sense in today's times.
Kind regards
Philipp Kern
that zVM. [...]
What about DIAGNOSE hypercalls to z/VM?
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
umented that "PREFETCH DATA (RELATIVE LONG)
signals the CPU to perform the specified operation, but it does not
guarantee that the CPU will necessarily honor the request.". So I
suppose a microcode update that disables the flushing is technically
fair
heir customers run
untrusted code got a heads-up. Also the scarcity of these machines helps
that researchers don't get around to test for the flaw on them. But
that's the usual security by obscurity defence[1].
Kind regards
Philipp Kern
[1] As Martha pointed out, apparently updates are in fact available.
that things keep running. They might experience more and more
breakage in the process.
Kind regards
Philipp Kern
Original Message
Subject: s390x ISA raised to z196
Date: 2018-02-18 19:17
From: Aurelien Jarno <aure...@debian.org>
To: debian-s...@lists.debian.org
Hi all,
Fol
zation generally doesn't nest either. It's possible but people
don't do it. Also partitioning a machine is not a thing.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
sen
s case VMware) about
this. There is something to be said for not leaking all of this information.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu w
the old Kerberos 4 server code that used to be part of VM TCPIP to current
> levels), and all the Linux distributions already support it.
Technically the acquired ticket is not two-factor, though. Instead it's
a bearer token that does not require reauth for
of Google products Go and v8 themselves got some porting
love, one for Docker and the other for Node.js.
Kind regards
Philipp Kern (not speaking for Google in any capacity)
--
For LINUX-390 subscribe / signoff / archive access instruc
But again you would not rely on
continuous availability of the OCSP responder in that case.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with t
ery clear design constraints around client systems.
Kind regards
Philipp Kern
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
ht
Hi,
Trying to increase visibility. Debian is looking for porters for s390x,
lest they will likely drop the port going forward. You don't need to be
a Debian contributor to help out, although it helps. [1] is the list for
coordination.
Kind regards and thanks
Philipp Kern
[1] https
74 matches
Mail list logo