We were just too busy to create a new release. However, a few of the Open Grid
Scheduler developers met back in June, and we are planning to release something
for SC14.
-Ron
--
On Tue, Aug 19, 2014 11:17 PM SGT Stu Whitman wrote:
Hello,
I have been using Open
Exit code 137 = process was killed because it exceeded the time limit, and
Google is your best friend if you have similar issues - and the solution is to
check the default time limit of your shell.
-Ron
Open Grid
There's also the SDM module that was released with SGE 6.2u5, and with SDM Grid
Engine can burst to EC2.
Not sure if anyone is still using SDM, but for those interested in learning
more about it:
https://blogs.oracle.com/templedf/entry/service_domain_manager
-Ron
- Original Message
Since we develop mainly Grid Engine and HPC code and know less about Apache
Cocoon, may be the Cocoon developers can give you some insight?
Is there really is more than one template named: use-qstatf? Or
the sitemap.xmap file has some invalid syntax in it?
-Ron
jfar...@uci.edu
To: Ron Chen ron_chen_...@yahoo.com
Cc: Rayson Ho rayray...@gmail.com; users@gridengine.org
users@gridengine.org
Sent: Monday, June 11, 2012 11:53 PM
Subject: Re: [gridengine users] PE Job Suspend / Resume
Thanks Ron for the details and explanation.
I will test NAMD
cgroup in the kernel was developed to help batch job management
systems, so if the system supports it, we should use it IMHO.
-Ron
- Original Message -
From: Joseph A. Farran jfar...@uci.edu
To: Ron Chen ron_chen_...@yahoo.com
Cc: Rayson Ho rayray...@gmail.com; users@gridengine.org
Can you compile that conftest.c program manually? If you are able to do that
without error, then make sure that you also use the same list of command line
arguments.
Then find out which option is needed to fail the compile. This way, you can
reduce the list of options needed to reproduce the
It's the java.lang.UnsupportedClassVersionError exception, most likely you are
using a higher version of JDK to build the source, but using a lower version of
JRE to run the generated class files.
You can google java.lang.UnsupportedClassVersionError and find out more about
this exception.
Can you check the generated config.log in the 3rdparty/qmake/WIN32_X86
directory? Usually, the last few lines of the config.log file tells you why the
compilation fails.
I haven't run the Windows SUA port of grid engine for a while. These days, for
client side only grid engine installations, I
...@istraresearch.com
To: Ron Chen ron_chen_...@yahoo.com
Cc: mahbube rustaee rust...@gmail.com; users@gridengine.org
users@gridengine.org
Sent: Sunday, June 10, 2012 1:33 AM
Subject: Re: [gridengine users] issue on build GE2011.11
Hi Ron,
Thanks for your quick feedback, however I am running
How about using the exclusive complex scheduling feature?
http://docs.oracle.com/cd/E24901_01/doc.62/e21978/management.htm#sthref431
And Reuti's reply to a similar
problem: http://gridengine.org/pipermail/users/2012-May/003535.html
-Ron
- Original Message -
From: Joseph Farran
In this case it is not a grid engine problem, but more of a X window problem.
Just google search X font path and you will get this
link: http://www.xfree86.org/current/fonts2.html
Then compare the output of xset q run under your account vs the other users
account.
$ xset q
-Ron
Hi Prakashan,
I am trying to understand your integration, and it looks like Ravi Chandra
Nallan's Hadoop Integration.
One of the improvements in Daniel Templeton's Hadoop Integration is he models
HDFS data as resources, and thus can schedule jobs to data. Is scheduling jobs
to data a planned
Ralph: How common will we see jobs that request dynamic allocations? I have
never seen Hadoop presentations talking about them in any BigData conferences.
Just also want to mention that Moab is not open-source, and I don't think we
will see much information about the integration from Moab.
William,
Oracle changed the Oracle Linux support structure in March this year. Quoting
the Senior VP of Linux and Virtualization Engineering, The nice thing is, if
you want a complete up to date system without
support, use this, if you then need support, get a support subscription.
Simple,
I think it is a great way to handle GPU permissions, and I sent your method of
managing GPU permissions to a GPU clustering discussion list.
However, I assumed that you were using multiple queue instances per node, and
each queue logically owns the GPU. Was my assumption correct?
-Ron
I found a few good papers on cgroups:
1. How I Used CGroups to Manage System Resources In Oracle Linux 6
http://www.oracle.com/technetwork/articles/servers-storage-admin/resource-controllers-linux-1506602.html
2. Introduction to Control Groups (Cgroups)
In RHEL 6 (and Oracle Linux 6, CentOS 6, or SL 6), there is the ns — the
namespace subsystem:
http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Resource_Management_Guide/ch01.html
However, the namespace subsystem was removed in recent kernels:
qstat only queries the qmaster for queued, running, or just finished jobs.
For jobs that are done, the qmaster writes the accounting information to the
accounting file:
http://gridscheduler.sourceforge.net/htmlman/htmlman5/accounting.html
And then qmaster clears the memory used by the finished
Note that we could not see the whole compile line, and sending images is not
the best way to let us know the issue.
Can you redirect the error to a file? You can do it by running aimk with
something like:
http://tldp.org/HOWTO/Bash-Prog-Intro-HOWTO-3.html
Then you can send us the last few
I agree with Hung-sheng.
If you are seeing NFS load issues, then you should switch to Local Spool Dirs.
It is very easy to use local spooling, just follow the guide
at: http://gridscheduler.sourceforge.net/howto/nfsreduce.html
While NFSv4 is better performing than earlier versions, you need to
Univa have New support for Postgres database job spooling balances speed of
submission with reliability in high volume clusters with lots of small jobs as
one of the new features of UGE 8.1 on their website.
We have discussed here before, PostgreSQL spooling was added by Sun, but was
taken out
Is it a typo? Wolfgang says there is only one Grid Engine Community community.
The 2nd community in the sentence is a dup.
-Ron
- Original Message -
From: Reuti re...@staff.uni-marburg.de
To: Miha Ahronovitz mij...@sbcglobal.net
Cc: users@gridengine.org; Miha Ahronovitz
I didn't even know the HPC Wire site before I read your e-mail.
Clicking on the URL redirects to univa.com, was there supposed to be content on
HPCwire?
-Ron
From: Prentice Bisbal prent...@ias.edu
To: users@gridengine.org Users users@gridengine.org
Sent:
I talked to an independent source recently, who told me that Grid Engine is not
safe. (He said Grid Engine as a whole is not legally safe).
So this is 6 months later, has the copyright infringement been fixed yet?
(I was away for a few months to handle a high priority contract, so I may have
And remember to add -soft so that it is biased but not treated as absolutely
essential:
http://gridscheduler.sourceforge.net/htmlman/htmlman1/qsub.html
-Ron
From: Simon Matthews simon.d.matth...@gmail.com
To: laot...@gmail.com
Cc: users@gridengine.org
If you are not using core binding, then Grid Engine does not allocate any
specific processors to the job.
It is up to the OS to assign processes (MPI tasks) to the physical processors.
-Ron
From: Ursula Winkler ursula.wink...@uni-graz.at
To: Reuti
From: Ursula Winkler ursula.wink...@uni-graz.at
To: Ron Chen ron_chen_...@yahoo.com; users@gridengine.org
users@gridengine.org
Sent: Wednesday, April 11, 2012 3:31 AM
Subject: Re: [gridengine users] double seizure of processors
Ron Chen wrote:
If you are not using core binding
Internally, all.q does not get any special treatment, and in fact you can
remove all.q from a grid engine cluster. So if the queue definitions are all
missing, then all.q should be one of them.
If your jobs were hung, you should find out why they were hanging. Using qstat
-f would be a first
The install script copies tcsh to qtcsh into the final installable package. So
you don't need to worry about not getting qtcsh.
-Ron
From: Winkler, Ursula (ursula.wink...@uni-graz.at)
ursula.wink...@uni-graz.at
To: 'Tim Wright' tim.dormo...@gmail.com
Cc:
I'm finally back from my 1-month Chinese New Year vacation (actually back since
Wednesday). A few things I wanted to add:
1) No matter what, Univa is not supposed to damage the reputation of Dave Love.
As much
as I disagree with Dave Love, I've never said bad things cowardly behind
someone.
FUD against open source, this is not the way to
revenge.
-Ron
- Original Message -
From: Ron Chen ron_chen_...@yahoo.com
To: Mark Magento markmaje...@gmail.com
Cc: users users@gridengine.org
Sent: Friday, January 6, 2012 12:46 AM
Subject: Re: [gridengine users] More Univa FUD???
Hi
Some limits are set in the job's environment, and there is no way to change it
once the job has started running already.
-Ron
From: Schmidt U. uschm...@mpi-halle.mpg.de
To: users@gridengine.org
Sent: Wednesday, January 11, 2012 1:42 AM
Subject: [gridengine
Hi Mark,
(Just back from my vacation and I am really late in this discussion.)
Did you create this website?
http://unicloud.wordpress.com/
While I am not a fan of Univa (mainly have problem with its market practices),
I am also not a fan
of those who create a website using the name of other
). We as a community should just move on. I will contribute a few
things to both the Open Grid Scheduler the Son of Grid Engine projects in
the future.
-Ron
- Original Message -
From: Dave Love d.l...@liverpool.ac.uk
To: Ron Chen ron_chen_...@yahoo.com
Cc: users@gridengine.org users
Chris,
1) There really are differences between Oracle Grid Engine and Univa Grid
Engine. First and foremost, Oracle has never used misleading or false
information just to get an extra customer to pay for Oracle Grid Engine. If you
have not read all the messages in the mail thread started by
mcheck is a GLIBC extension. So commenting out those lines are fine, but when
we integrate the changes to the source tree, it is better to check for the
__GNU_LIBRARY__ macro.
http://www.gnu.org/s/hello/manual/gnulib/mcheck_002eh.html
-Ron
From: Rayson Ho
Dave,
At the beginning of this year, I warned both you and Rayson (who was already
the owner of the Open Grid Scheduler project - as I was busy with my personal
commitments) to be aware of a company that never contributed anything to SGE
and suddently poped up and wanted to be the owner of the
--- On Fri, 9/9/11, Mark Dixon m.c.di...@leeds.ac.uk wrote:
I'm still a bit fuzzy on what differentiates the efforts,
and am concerned that the projects might not be pulling in a
common direction to benefit the community as a whole.
Mark,
In fact our existance benefits the community -- did
:40, Ron Chen ron_chen_...@yahoo.com
wrote:
--- On Fri, 9/9/11, Mark Dixon m.c.di...@leeds.ac.uk
wrote:
I'm still a bit fuzzy on what differentiates
the efforts,
and am concerned that the projects might not
be pulling in a
common direction to benefit the community as a
whole
and
copyright reassignment
Sent from my iPad
Hung-Sheng Tsao ( LaoTsao) Ph.D
On Sep 9, 2011, at 12:21, Ron Chen ron_chen_...@yahoo.com
wrote:
--- On Sat, 9/10/11, William Bryce bbr...@univa.com
wrote:
Hi Ron,
I suggest you take your 'tin foil hat off'
Ron.
Hmm, so you
--- On Tue, 8/16/11, William Bryce bbr...@univa.com wrote:
I would love to be making lots of money on Grid Engine :-)
but not surprisingly it costs $$ to hire developers, write
documentation and develop new functionality in the product.
Regards,
Bill.
On 2011-08-15, at 12:00 PM, Ron Chen
As Univa is selling Grid Engine for money, making lots of money and expanding,
I hope Univa would contribute back to Grid Engine, not only in the form of code
changes (which are really bug fixes so far), but something bigger - like
contribute the documentation to open-source -- so far everyone
Andreas,
I just worked with Rayson and got it working for 32-bit -- I will check it in
soon.
-Ron
--- On Tue, 6/21/11, Rayson Ho rayray...@gmail.com wrote:
From: Rayson Ho rayray...@gmail.com
Subject: Re: [gridengine users] OGS build problem on SL5.5 i386
To: Andreas Haupt
Hi Kunio,
So you are saying that only SGE is upgraded and then X forwarding does not work?
From the error messages, I believe it has to do with something else, as the X
client is trying to open the display at localhost:11.0 on the execution host!
From SGE point of view, it just executes the
--- On Sat, 3/19/11, Stephen Dennis sden...@univa.com wrote:
So, if user has 14 group memberships the the add of the
supplemental gid will fail and be treated as an error.
For SGE 6.2u5 the work around is to ensure that you have
fewer than 14 group membership for grid engine users.
This is
Hi Dave,
I would like to see if we can do some inter-fork collaboration here on the list.
Can you review the fix for this issue? If you think it is OK, then I will check
it into the Open Grid Scheduler project, and you are free to check it into the
Son of Grid Engine project.
Note that I
If all you want is a simple migration from PLPA to hwloc, then yes it is
fairly straightforward. But then, you don't get the advantage of using hwloc,
all you get is the extra OSes support.
If you want to get the extra information about NUMA nodes, for example, then
you will need to add extra
--- On Wed, 3/16/11, Dave Love d.l...@liverpool.ac.uk wrote:
I was assuming the private branches that have been merged
were just for
development, which is the right way to work. I didn't
mean you were keeping anything back.
While we are going to talk about merging the forks in the next Grid
Lane,
I hope you understand that if you don't provide enough info, then we can't
really help you.
Without knowing the context, and without backgroud info about the cluster load,
etc, I don't know how much longer we need to guess why SGE is not working.
Is it really that hard to gather post
50 matches
Mail list logo