S
Having had a (several) play with the scheduler toy, the only thing I can
confidently say is I'm not much better placed to help. I have no time slots
without a conflict of some sort.
I (currently) have no slots in the Linux/VM stream where I will definitely be
attending, despite there
I was never much enamoured with the original cpuplugd - and my customer
response to testing it was less than enthusiastic. This, however, looks more
interesting.
Now to see if I can convince the customer to retest using it (R/H 6.4).
Shane ...
On Fri, Jun 28th, 2013 at 1:12 AM, Bruce Hayden
On Sat, Jun 15th, 2013 at 7:52 AM, Jagos, Brian V wrote:
Hi All,
It is that time of year again this is a call to chair SHARE
session.
Hadn't realised the schedule was out.
Ugh - now I remember why the various streams have so much trouble getting
volunteers for chairs. It's
And just for completeness, I have a customer happily running 5.4 on a z114.
I would go for 6.2 on the new box, but you'd have to think IBM would cane you
something awful with licensing for running z/OS as guests.
Needs must I suppose ...
Shane ...
On Tue, May 28th, 2013 at 8:09 PM, Feller, Paul
I'm sure this will induce Philipp Kern to rise to the task.
However a quick search on this list will also get you Fedora - that might
suffice for educational purposes. Especially if you are RHEL inclined.
CentOS used to do a s390x build, but I haven't seen that in years.
Shane ...
On Fri, May
On Fri, May 10th, 2013 at 2:32 AM, David Boyes wrote:
You're probably not going to budge them on that.
With luck maybe Filipe can bring some more clout to the table ;-)
Shane ...
--
For LINUX-390 subscribe / signoff /
On Thu, Mar 21st, 2013 at 4:02 AM, Ben Duncan wrote:
*SNARK* .. Gave up Java for Python.
lol - I went the other way. No real loss in my case as my Python was
rudimentary, and I wanted to write an app for my phone
Now everyone tells me I should be on HTML5 d'oh.
Shane ...
On Thu, Jan 31st, 2013 at 12:44 AM, John McKown wrote:
Thanks to all for the input! I _tried_ to run the script over night. I
added an echo to tell me which input file I was working on. I came in
this morning. It had been running from 14:00 to 06:30 (16 1/2 hours)
and was still on the first
I've had no trouble reading your mails from a (linux) mail client or a web
interface to my ISP mail q.
Using a web interface to linux-390, your messages disappear, leaving just
attachments.
Shane ...
On Tue, Jan 29th, 2013 at 2:59 AM, Richard Troth wrote:
friends --
Sincere apologies for
On Thu, Jan 10th, 2013 at 10:59 AM, Rick Troth wrote:
GPT is s conspiracy of the partition police and the UEFI underworld.
A little paranoia is good for the soul ... :-)
Partition tables are needed ... sometimes ... not always. (One case
where they are needed is when GRUB and UEFI gang
On Fri, Dec 21st, 2012 at 12:40 PM, Jagos, Brian V wrote:
Yes SHARE is right around the
corner and it is time for a call for CHAIRS.
I am constantly baffled that the various streams appear to have so much
trouble getting people to help on this. I can vouch that it ain't a tough
gig, and you
There is a thread on another list re why I love my z vendors.
Clear evidence here.
Excellent, pure and simple.
Shane ...
On Tue, Dec 4th, 2012 at 6:34 PM, Dan Horák wrote:
yes, I've checked the sources for dracut (the tool that downloads the
image after setting the needed devices online) and
Mark Post (from Suse) spake thus:
RHEL ships with tigervnc (the client) and tigervnc-server. Is there some
problem with those?
Who says they don't keep an eye on their contemporaries ... :0)
Keep up the good work, one and all.
Shane ...
Might explain the quality of the doco - kudos as appropriate.
Shane ...
On Sat, Sep 1st, 2012 at 10:07 PM, Thang Pham wrote:
As of July 13, 2012, a service contract for xCAT on z/VM is available for
purchase from IBM. More info is available on
http://www.vm.ibm.com/sysman/
As per the document one can modify the the syslog-ng.conf file to supress
the messages.
React to the symptom (and hide the evidence) rather than fix the actual problem.
Has happened before, will do again.
Shane ...
--
For
On Sat, Sep 1st, 2012 at 1:17 AM, Lee Stewart wrote:
Is anyone using xcat (http://xcat.sourceforge.net/) in production? Any
comments on it?
Nope, but thanks for asking the question. Looks interesting.
My only comment would be that the z{VM,Linux} doco looks extremely well done.
Shane ...
And ... ?.
http://gigaom.com/cloud/some-of-amazon-web-services-are-down-again/
Shane ...
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390
From a non-z perspective it is excellent. It's all (??? - the bits I've
looked at anyway) perl, has a daemon mode that creates a detailed history,
and is fine-grained in the data.
Mark, the developer, is very receptive to suggestions.
Must see if I can get it on a (z) customer site somewhere.
Seems this is a common refrain in this neck of the woods.
Lots of blue-sky stories, then little or no engagement.
There are (significant) successes, of which there are a few people subscribed
here, but there are also a number that are (deliberately) unpublicised.
Very unfortunate.
zLinux in Aus
On Tue, Jul 17th, 2012 at 6:37 AM, Alan Altmark wrote:
...
TCPIP is connected to two HiperSocket networks: one real (to MVS) and one
virtual (to Linux).
Now, hold it right there fella. I want the order number for one of those
*real* hipersockets.
(haven't we been here before ... ;-)
Shane ...
It should be noted top accepts a parm to adjust the report frequency - all
the way down to fractions of a second. *All* sampling based monitors have
their weaknesses - especially those that run in userspace.
However, for those of us that developed our performance tuning/debugging
skills in a
Be careful what you ask for.
This is _not_ a true (time-of-day) timestamp. At least it wasn't last I
looked. Think kernel active time since boot - useful for relative (timed)
occurrences for kernel events.
It would be reasonably trivial to adjust it to a ToD stamp, but it's going to
have some
On Fri, Apr 13th, 2012 at 3:41 PM, Mark Post wrote:
/proc/*/smaps exists in SLES10, but there's no Swap: fields in them.
Arrgggh. That'll certainly get a bunch of zeroes out of that script.
Note to self, _never_ presume nuthin ...
Thanks Mark.
No, by lazy in this context I meant that freed memory (pages) are not
immediately moved to the free list. This even extends to task termination.
If memory pressure ramps up sufficiently, kswapd will get kicked to balance
out the trees. Could take a while - like forever.
In addition to what Rob
On Thu, Apr 12th, 2012 at 8:37 PM, Rob van der Heij wrote:
With 22 processes having this mapped, I would count it as 22 times 8
kB while it really is just 8 kB on swap? And how come part of this is
private when it's read-only?
Note the last sentence of my previous post. That applies
- obvious first step would be to check the entire system, rather than a
subset you obviously think is the cause.
- smaps should be believed.
- Linux uses lazy (memory) allocation. This includes de-allocation. And swap.
Hence the various tools that simply read meminfo should be treated with
On Thu, Apr 5th, 2012 at 6:06 PM, Lu GL Gao wrote:
We know that linux cpu usage mainly include user cpu and sys cpu. But why
performance toolkit value cannot corresponding with top command value?
heresy
Why not indeed.
Hipervisors are becoming a commodity item. IBM (and its ISVs) has fought
Sorry Alan - I trust your paroxysm of coughing did no lasting damage ;-)
I meant the concept, not necessarily that precise implementation.
Shane ...
On Sat, Apr 7th, 2012 at 3:22 AM, Alan Altmark wrote:
(cough) With VIF, we learned an important lesson on how NOT to make a
hypervisor.
And is already linked off the sourceforge page. Excellent work !.
Shane ...
On Wed, Mar 21st, 2012 at 9:25 PM, Dorothea Matthaeus wrote:
The Linux Health Checker User's Guide will soon be available also on
developerWorks.
C'mon Mark, be a bit more positive lol.
Let's hope Barbara doesn't get to hear of this (*) ... ;-)
Shane ...
(*) - apologies to those not on IBM-MAIN
On Thu, Mar 22nd, 2012 at 2:25 AM, Mark Post wrote:
I'm more interested in a mechanism to report bugs. From perusing the
SourceForge
It doesn't matter much - all the important meta-data is at the front and
gets clobbered first.
Usually.
LVM is greatly lauded as an Enterprise solution, but the initial design was
terribly flawed, and numerous iterations to rectify it have not been entirely
fruitful.
IMHO of course.
Even if only
This is just your friendly vendor messing with you.
Not to mention LVM itself.
An lv is not a volume, it's really a partition - no, wait, that's what a pv
is ...
Unless, of course, a pv is a full volume, and not a partition at all.
And a vg is a group of volumes except when a pv is not a
C'mon Rodger, stop beating around the bush - say what you really mean ;-)
Whenever I've sat in on any presentations by database developers , they
*always* want all the memory, to do their own (direct) I/O, avoid O/S
services (like caching) and to hell with the rest of the users of system
I must admit some ambivalence to being hamstrung by such standards.
I did enjoy having a read of this:
http://lists.busybox.net/pipermail/busybox/2010-December/074114.html
Shane ...
...
The default in /opt/IBM is not a great option.
Per the FHS, I believe it should have been (should be)
On Fri, Feb 10th, 2012 at 11:00 PM, Joerg Reuter wrote:
On Thu, Feb 09, 2012 at 04:43:25PM -0600, Ron Foster at Baldor-IS wrote:
The only messages I can find have to do with a hipersockets time out.
Feb 9 11:21:39 bus0104 kernel: NETDEV WATCHDOG: hsi0: transmit timed
out
Feb 9
On Wed, Nov 9th, 2011 at 3:44 PM, Alan Altmark wrote:
Just because the current crop of security weebles don't get it does
not a true problem make.
Eh?
LOL.
Should it come to pass that Alan and I are once again in the same bar
imbibing the best of Aussie brews/wine, I must remember to
Have you considered alternatives - pax for example ?.
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
On Thu, Jun 9th, 2011 at 6:12 AM, Mark Post wrote:
To be honest, I don't think LVM itself cares about the partition type.
pvcreate will use any partition you point it at.
This is unfortunately (almost unbelievably) true - with the notable exception
of swap. Regardless of the partition type of
H - and why isn't this generally available ?.
Open source ... ???
Not meaning to piss on the messenger, but I'm happy to do likewise to the
policy makers involved.
This is just *bad*.
Put it out there, let the community benefit.
Having been the recipient of IBMs progression to OCO, this just
... which allows write access to a CMS filesystem.
Note the *write* access - is this generally available ?.
Shane ...
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
Hard to argue ... lol
Shane ...
On Fri, Apr 1st, 2011 at 10:15 PM, Mauro Souza wrote:
But anything is better than twm.
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
And which real address 0 might that be ?.
Remembering that most people will be running as a guest under a hipervisor
(z/VM) running second level under another hipervisor (PR/SM).
I note our German maintainers have been conspicuously quiet
Shane ...
On Thu, Mar 17th, 2011 at 6:32 AM, Mark Post
As Rob pointed out, good luck trying to figure what that actually resolves to
hardware-wise.
Shane ...
On Thu, Mar 17th, 2011 at 8:46 PM, Heiko Carstens wrote:
The kernel gets loaded to address absolute zero ...
--
For
And another good thread was had by all.
Whilst I suspect the initial question was in no way related to the problem
observed, the asking elucidated some fine information on the zSeries port for
those of us with a morbid fascination in such matters.
Thanks to all.
Shane ...
I've been known to drop files in /tmp for later collection - by myself or
others.
Have you considered skulker ?.
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the
Just to clarify, this was based on the OE reference - i.e Unix Systems
Services running under z/OS rather than zLinux.
Shane ...
On Sat, Mar 12th, 2011 at 1:44 AM, I wrote:
Have you considered skulker ?.
--
For LINUX-390
My perspective on this is that swap is there to soak up allocations you
weren't prepared for.
Who cares about the cost of (virtual) disk allocation.
As for Rob (who works for someone that sells software monitors) beating up on
someone who works for the hardware vendor ... ???
Chill fella ... just
You wish
You'll find them scattered hither and yon - especiallly with LVM faking
a(nother) block device layer.
And, as you've already discovered, the second U is a lie. Sometimes ...
Shane ...
On Wed, Feb 2nd, 2011 at 10:48 AM, Alan Altmark wrote:
Ah, so these UUIDs are not the builtin
And as a dispassionate (database ignorant) observer, I'm happy to see such a
response from a vendor.
Top points.
Shane ...
On Wed, Dec 22nd, 2010 at 10:17 PM, Damian Gallagher wrote:
The Oracle DBA indicates that Oracle is easier to implement under Red
Hat - I'm interested in what this
Who is this Altmark fella anyway ?.
Shane g,d,r
On Thu, Sep 9th, 2010 at 10:24 PM, Alan Altmark wrote:
...
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu
Whilst I'm sure lots of people (in the northern hemisphere) appreciate the
running of this twice in the one day, this still equated to 23:00 and 04:00
for us on the Australian east coast. Much as I'd like to partake ...
The ability to later re-run the session(s) at my leisure is much appreciated.
Hmmm - high sys CPU usage, high loadavg, system not talking to anyone.
Smells like it's busy doing its own stuff. If it were me I'd want to know
trends for things like
swap-in and swap-out rates, tasks in uninterruptible sleep, context switch
counts.
SAR is too granular to be any use even if
Short answer, no.
This was discussed earlier in the year - see:
http://www.mail-archive.com/linux-390@vm.marist.edu/msg55911.html
Shane ...
On Tue, Jul 27th, 2010 at 1:17 AM, Marcy Cortes wrote:
Is there a way to limit the amount of memory used for page cache?
Some more info please.
... you get a OOM condition ?.
... the/a large consumer gets killed ?
... the system halts (explain) ?.
You *want* a system-wide panic ?. If so, setting /proc/sys/vm/panic_on_oom to
1 will
have the desired effect on non zLinux.
Shane ...
On Tue, Jul 27th, 2010 at 1:24
This is somewhat empirical, and depending on the usage of shared libraries
will over-estimate the situation somewhat.
Not a bad thing in the context.
Determining actual accountable memory on linux has been like catching a
greased pig. With recent kernels you can work back from the pagemaps -
I would be guessing you have lots of Oracle threads - all of which will have
the same common code mapped. And counted by the summation.
These fields only count resident memory - swap usage will be (sort of)
irrelevant. Pages that are swap cached (as distinct from only swapped out)
reside in real
Sorry ... that should be compressed _swap_ cache
Shane ...
On Sat, Jul 24th, 2010 at 9:00 AM, I wrote:
... Recently the memory manager has
started using compressed page cache to ameliorate this effect somewhat.
--
For
On Sat, Jul 24th, 2010 at 10:20 AM, Marcy Cortes wrote:
Does this imply that the best setting for Linux on z is to use the FILE
SYSTEM CACHING (Direct i/o disabled)?
I won't presume to be able to answer that, but I will observe that Linus has
made some very harsh comments about database
Barton yanking Alans chain ...
Where have I seen that before ?.
Shane ...
On Fri, Jul 23rd, 2010 at 5:26 AM, Barton Robinson wrote:
Alan, are you trying to make this announcement so totally boring on
purpose? Just business as usual? nothing really new and exciting? Is
there anything here
All the bits are now to be OSA connected.
Shane ...
On Fri, Jul 23rd, 2010 at 8:40 AM, Mark Post wrote:
I didn't see any mention of an IP network, just that it was private.
That could mean a lot of things.
--
For LINUX-390
I might just add that despite it's manpage assertion, rsync isn't too
intelligent about it at all.
My (non z) testing indicated that if you re-use the same target file, after
the initial run cp is significantly more efficient. The initial run for
both is comparable as the target needs to be
Seems Jim couldn't keep the lid on things.
The IBM Canada home page is all dressed up with nowhere to go ... :0)
Shane ...
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to
Maybe you should be thankful Mark didn't go with the default RoT of
swap=2xRAM
Personally I've never allowed _any_ installer to handle partitioning,
but I can appreciate your concern for the novice installer (person).
They will/should probably be using mod-9s anyway, rather than trying to
squeeze
Maybe - but a hacker wouldn't.
But I'll bet he keeps his trap shut in future.
Shane ...
On Wed, Jun 9th, 2010 at 11:38 PM, Mrohs, Ray wrote:
What we really just saw is another young person being turned away
from z/VM, at a time when the next generation needs to be engaged and
involved
Just to be clear, I use hacker in this sense:
http://en.wikipedia.org/wiki/Hacker_(programmer_subculture)
Shane ...
On Wed, Jun 9th, 2010 at 11:56 PM, I wrote:
Maybe - but a hacker wouldn't.
--
For LINUX-390 subscribe /
On Sun, Jun 6th, 2010 at 2:13 AM, Alan Altmark wrote:
... a great candidate for a wiki article. There's even a placeholder just
waiting for you Alan ... :0)
This really is such a can of worms it needs something authoritative out where
people can easily find it.
Shane ...
C'mon Alan, enough of the equivocation.
Try and give us at least a semblance of what you *really* think about this ...
;-)
Shane ...
On Wed, Jun 2nd, 2010 at 6:49 AM, Alan Altmark wrote:
Multinetting is a Bad Idea, to be stomped out of existence and never
mentioned again in polite company.
I am constantly amazed when I (still) see references to LKCD. Is this the only
environment still using it in
preference to Redhats crash ?.
I dropped off the tech mailing list *years* ago when it died. At one point I
was looking at contributing to
lcrash, but gave it away as a pointless
Were it me I'd be thinking idle workload(s) having its/their storage stolen.
That implies some (heavy) overnight other I/O workload - backups, updatedb, ...
Swappiness was invented for just this scenario - what is yours set at ?.
Top and the ilk won't be of much help - a large (swap) I/O spike
Meandering around from said link I found myself at the (IBM) vm packages
download page.
Top entry was CP3KVMXT.
In my caffeine deprived state first thing in the morning all I saw was ...KVMXT.
Mmmm, thinks I.
Followed immediately by ... damn.
Shane ...
Have you tried going to the source?
Now there's a lad that gets around.
He was speaking here in Aus just a couple of weeks back.
Shane ...
For information, the details of the next meeting of the UK zLinux
User Group:
...
15:20 - 16:00 Accounting and Chargeback
Barton
Be aware that Mark answered the question you asked.
This will not necessarily be the same should you choose to change distro
sometime in the future.
Shane ...
On Thu, May 13th, 2010 at 9:41 AM, Billy Bingham wrote:
Thanks Mark. I'm coming from a VSE background and this is taking some
getting
Hopefully doing its job - servicing other guests. What are you asking ?.
Linux metrics are from the perspective of Linux - and it has a heritage of a
stand-alone O/S. It may
think the cycles have been stolen, others may disagree.
Think of it as involuntary non-dispatch. Has to happen in any non
Hmmm - Red on Green ... nah.
Maybe Red on Red ... :)
Some-one's gotta have a picture of a nice red Amdahl box somewhere ...
Of course, they ran Unix (UTS) rather than Linux - Oh well, scrap that idea I
suppose.
Shane ...
On Fri, Apr 30th, 2010 at 7:33 AM, Hodge, Robert L wrote:
I would be
On Thu, Apr 8th, 2010 at 4:58 AM, Ron Foster wrote:
The DBAs cancelled the DB2 thread. The SAP Administrators issued all the
commands thru
SAP to get rid of the process. They did not work. The process did not
go away. Then we tried a kill -9 and the process died.
On the other two
I usually just open another (virtual) terminal and kill the ping.
Sounds like you might benefit from defining an alias for ping.
Shane ...
On Mon, Apr 5th, 2010 at 9:26 AM, Clovis Pereira wrote:
Hi,
Sometimes, I forgot to type the cont parameter using the ping command
on a
3215 terminal
Generally I'm with Christian on this one.
The procedure Mark linked is appropriate for (most) *sub*-directories of
the the root - not the entire root itself. The pseudo directories need
to be excluded, and things like logs, even in single user, are
problematic - that may not be an issue on test
How disappointing ... :0)
I was another thinking it was just Java being Java.
Shane ...
On Fri, Mar 12th, 2010 at 8:42 PM, Martin Schwidefsky wrote:
MVCLE with a zero source length is a memset. The instruction only
accesses the target address, the source address is of no concern.
A
I had fired up a torrent to see if it would help Bernie's cause, but
several hours later there had been no traffic - in or out. Due to noise
(overnight at home) and power usage, I killed the box.
This is the flaw in relying on torrents for low usage sources.
Will auditors allow it at work ?? ...
On Thu, Mar 4th, 2010 at 2:49 PM, Bernie VK2KAD wrote:
I am trying to download Centos for s390x - I think the latest is 4.7
Unfortunately, the only download for a DVD ISO image seems to be via
bittorrent - P2P software is a no-no in my organisation so I have had
a
torrent client trying to
The umount *has* to complete, even if the f/s is broken - else you can't fix
it. Well, you could try,
but let's not go there.
Of course, what a fsck treats as fixed and what we as users might like to
think it produces may
be totally at odds. Truncating files and unlinking inodes ain't my idea
81 matches
Mail list logo
