mauischeduler-users Mailing List for Maui Scheduler
Brought to you by:
mauisched,
mitchmurphy
You can subscribe to this list here.
| 2000 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
(11) |
Dec
(10) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2001 |
Jan
(16) |
Feb
(26) |
Mar
(28) |
Apr
(6) |
May
(5) |
Jun
(2) |
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
(1) |
| 2002 |
Jan
|
Feb
(2) |
Mar
|
Apr
(23) |
May
(7) |
Jun
(5) |
Jul
(3) |
Aug
(6) |
Sep
(10) |
Oct
(13) |
Nov
(2) |
Dec
(2) |
| 2003 |
Jan
(5) |
Feb
(5) |
Mar
(2) |
Apr
(9) |
May
(9) |
Jun
(9) |
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(2) |
| 2004 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
(2) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(3) |
Jun
|
Jul
(1) |
Aug
(1) |
Sep
(1) |
Oct
|
Nov
|
Dec
|
| 2006 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2007 |
Jan
(4) |
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
(1) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2012 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
| 2013 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
| 2016 |
Jan
|
Feb
(1) |
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Monica M. <mar...@gm...> - 2016-04-11 16:33:51
|
Hello, I am running Maui in Simulation mode with the parameter SIMSTOPITERATION set to 2000. However, the simulator exits after 9 iterations, and does not queue the jobs for which resources are not available in the 1st iteration. Can anyone suggest a solution for this? I am using maui-3.3.1 Thanks, Monica |
|
From: Monica M. <mar...@gm...> - 2016-02-16 03:28:44
|
Hello, I am trying to understand the Maui source code and it's flow when used with Torque resource manager. Is there any documentation to help with this? I need to study the source code and then add a specific scheduling policy to it. Thanks, Monica -- - Monica Marathe |
|
From: joes_ith <joe...@zo...> - 2013-10-06 10:08:25
|
http://www.embl.de/aboutus/jobs/searchjobs/index.php?newlang=1&ref=EBI_00329&back=%2Faboutus%2Fjobs%2Fsearchjobs%2Findex.php%3Floc%3D2%26list%3D1 |
|
From: Anista R. <icm...@ic...> - 2012-11-06 12:34:15
|
Dear all, a job that has finished some hours ago is still running in the batch queue, how can it be possible? some mistake in the configuration or a user mistake? I can't understand what is happening, any suggestion? Thanks in advanced |
|
From: Ben A. <ben...@st...> - 2007-06-25 09:05:24
|
Hello there I would greatly appreciate a small amount of your time to assist with my doctoral research at The University of Newcastle. The research concerns open source licensing and we're seeking developers working on Java projects. The research is supervised, ethics-approved, anonymous and results will be freely available. Participation will also provide a custom licensing report for your project. To learn more, please visit: http://licensing-research.newcastle.edu.au Thanks for reading this email, and I hope you'll consider participating. Best regards Ben Alex (My apologies for being off-topic; this list will not be emailed again) |
|
From: Glen B. <gle...@gm...> - 2007-02-06 20:18:01
|
http://www.clusterresources.com/pages/products/maui-cluster-scheduler.php |
|
From: Roger D. <rog...@gl...> - 2007-01-09 13:52:39
|
Is maui still alive? Maui seems to be dead which is a shame as Maui appears to be the only OSS enterprise scheduler out there. I remained on this list by accident. Cheers Roger At 09/01/2007 13:23:10, mau...@li... wrote: >Hi, > >I'am just start to work with the maui scheduler an have some questions how >I can get some more detail informations from maui. >Is it posible to get the current priority of a job in the queue? Or are >there a interface like python or perl bindings to access internal data? > >I want to get this information to check if the scheduler is well >configured. So I can check which user got a to high/ to low priority. > >thanks a lot >Clemens > >-- >Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ > > >------------------------------------------------------------------------- >Take Surveys. Earn Cash. Influence the Future of IT >Join SourceForge.net's Techsay panel and you'll get the chance to share your >opinions on IT & business topics through brief surveys - and earn cash >http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV >_______________________________________________ >Mauischeduler-users mailing list >Mau...@li... >https://lists.sourceforge.net/lists/listinfo/mauischeduler-users > |
|
From: Ole H. N. <Ole...@fy...> - 2007-01-09 13:52:13
|
Clemens Zeidler wrote: > I'am just start to work with the maui scheduler an have some questions how > I can get some more detail informations from maui. > Is it posible to get the current priority of a job in the queue? Or are > there a interface like python or perl bindings to access internal data? > > I want to get this information to check if the scheduler is well > configured. So I can check which user got a to high/ to low priority. 1. showq 2. diagnose -p 3. diagnose -f Best regards, Ole -- Ole Holm Nielsen Department of Physics, Technical University of Denmark |
|
From: Clemens Z. <cle...@rw...> - 2007-01-09 13:49:44
|
Hi, I'am just start to work with the maui scheduler an have some questions how I can get some more detail informations from maui. Is it posible to get the current priority of a job in the queue? Or are there a interface like python or perl bindings to access internal data? I want to get this information to check if the scheduler is well configured. So I can check which user got a to high/ to low priority. thanks a lot Clemens -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ |
|
From: Clemens Z. <cle...@rw...> - 2007-01-09 13:26:07
|
Hi, I'am just start to work with the maui scheduler an have some questions how I can get some more detail informations from maui. Is it posible to get the current priority of a job in the queue? Or are there a interface like python or perl bindings to access internal data? I want to get this information to check if the scheduler is well configured. So I can check which user got a to high/ to low priority. thanks a lot Clemens -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ |
|
From: Atwood, R. C <r.a...@im...> - 2006-02-14 15:12:02
|
I believe I unscubscribed from this list a year or so ago, are old addresses getting put back in by some ghost or backup restore or something? 'Cause this is the first email for this list that I have received in all that time. Just thought I should mention such an odd event in case it indicates a problem for the list Anyways I will try to unsubscribe again..=20 -----Original Message----- From: mau...@li... [mailto:mau...@li...] On Behalf Of David Oliveira Sent: 14 February 2006 15:07 To: mau...@li... Subject: [Mauischeduler-users] unm.maui.misc.logger Importance: High |
|
From: David O. <dbo...@st...> - 2006-02-14 15:03:50
|
Hi All,
The class Logger under unm.maui.misc contains the rotateLog procedure that
rotate a specific Log file. However sometimes is desirable to rotate log
files and maintain the inode in order to maintain the stream and the file
descriptor. If we have an arbitrary application writing to a file that
will be rotated in the next second with this class, the application will
lose the ability of writing to that file because there was an inode
change.
public static OutputStream rotateLog (
...
// change back
// remove the old file
logFile.delete();
return new FileOutputStream(logFileName);
}
Comment the logFile.delete() line will "fix" this issue, probably this is
not a fix because the current code can perform the desiable functionality,
however, I think you should know this(you already know... but well...)
Gracias,
David O.
|
|
From: Clarence D. <di...@sa...> - 2005-09-19 00:31:24
|
I have dual cpu nodes on my cluster and I'd like to configure maui to assign nodes to newly submitted jobs such that the nodes haven't been completely filled with jobs yet. The scheduler as it is configured with NODEALLOCATIONPOLICY PRIORITY NODECFG[DEFAULT] PRIORITYF=JOBCOUNT still does the same as it did out of the box. It assigns new nodes rather than filling up existing nodes (see the qsub script below). Thus, let's say I have node1/0, node1/1, node2/0, node2/1, node3/0, node3/1, etc. Under the current configuration, maui just stupidly assigns node1/0, node2/0, node3/0, etc, rather than node1/0, node1/1, node2/0, node2/1, node3/0, node3/1, etc. How should I correctly configure maui to follow this sequential assignment order? Here is an example qsub script. ---- qsub script start ---- #!/bin/tcsh #PBS -l walltime=1:00:00 #PBS -l mem=400mb #PBS -l nodes=1 # How many nodes did we ask for? set NN = `cat $PBS_NODEFILE | wc -l` lamboot /usr/local/lam-7.1.1-64bit-gnu/bin/mpirun -np 1 /home/wsm/bench/pi lamclean lamhalt ---- qsub script end ---- For completeness, I have attached the entire maui.cfg file at the end of this message. -- = Clarence K. Din = Director of Computer Facilities / Dept of Chemistry / Univ of Pennsylvania = office: CHEM B-12 / email: di...@sa... / ph: 215 898 4714 / fx: 215 573 2112 -- Today's fortune... You will have good luck and overcome many hardships. maui.cfg # maui.cfg 3.2p8 SERVERHOST head # primary admin must be first in list ADMIN1 root # Resource Manager Definition RMCFG[HEAD] TYPE=PBS # Allocation Manager Definition AMCFG[bank] TYPE=NONE # full parameter docs at http://supercluster.org/mauidocs/ a.fparameters.html # use the 'schedctl -l' command to display current configuration RMPOLLINTERVAL 00:00:20 SERVERPORT 42559 SERVERMODE NORMAL # Admin: http://supercluster.org/mauidocs/a.esecurity.html LOGFILE /usr/local/maui-3.2.6p11/log/maui.log LOGFILEMAXSIZE 10000000 LOGLEVEL 3 # Job Priority: http://supercluster.org/mauidocs/ 5.1jobprioritization.html QUEUETIMEWEIGHT 0 # FairShare: http://supercluster.org/mauidocs/6.3fairshare.html FSPOLICY PSDEDICATED FSDEPTH 7 FSINTERVAL 86400 FSDECAY 0.50 FSWEIGHT 100 FSUSERWEIGHT 25 # Throttling Policies: http://supercluster.org/mauidocs/ 6.2throttlingpolicies.ht ml # NONE SPECIFIED # Backfill: http://supercluster.org/mauidocs/8.2backfill.html BACKFILLPOLICY FIRSTFIT RESERVATIONPOLICY CURRENTHIGHEST # Node Allocation: http://supercluster.org/mauidocs/ 5.2nodeallocation.html # NODEALLOCATIONPOLICY changed, hurtubis 7/27/05 NODEALLOCATIONPOLICY PRIORITY NODECFG[DEFAULT] PRIORITYF=JOBCOUNT # Added by WSM JOBNODEMATCHPOLICY EXACTNODE NODEACCESSPOLICY SHARED JOBMAXSTARTTIME 00:00:20 DEFERTIME 0 JOBMAXOVERRUN 0 # # QOS: http://supercluster.org/mauidocs/7.3qos.html # QOSCFG[hi] PRIORITY=100 XFTARGET=100 FLAGS=PREEMPTOR:IGNMAXJOB # QOSCFG[low] PRIORITY=-1000 FLAGS=PREEMPTEE # Standing Reservations: http://supercluster.org/mauidocs/ 7.1.3standingreservati ons.html # SRSTARTTIME[test] 8:00:00 # SRENDTIME[test] 17:00:00 # SRDAYS[test] MON TUE WED THU FRI # SRTASKCOUNT[test] 20 # SRMAXTIME[test] 0:30:00 # Creds: http://supercluster.org/mauidocs/6.1fairnessoverview.html # USERCFG[DEFAULT] FSTARGET=25.0 # USERCFG[john] PRIORITY=100 FSTARGET=10.0- # GROUPCFG[staff] PRIORITY=1000 QLIST=hi:low QDEF=hi # CLASSCFG[batch] FLAGS=PREEMPTEE # CLASSCFG[interactive] FLAGS=PREEMPTOR # USERCFG[DEFAULT] FSTARGET=5.0+ MAXPROC=44 USERCFG[DEFAULT] FSTARGET=100.0+ MAXPROC=44 # advanced users # USERCFG[josh] FSTARGET=45.0+ MAXPROC=44 # USERCFG[per] FSTARGET=45.0+ MAXPROC=44 USERCFG[josh] FSTARGET=100.0+ MAXPROC=44 USERCFG[per] FSTARGET=100.0+ MAXPROC=44 USERCFG[wsm] FSTARGET=100.0+ MAXPROC=44 |
|
From: Clarence D. <di...@sa...> - 2005-08-18 17:52:27
|
I have dual cpu nodes on my cluster and I'd like to configure maui to assign nodes to newly submitted jobs such that the nodes haven't been completely filled with jobs yet. The scheduler as it is configured with NODEALLOCATIONPOLICY PRIORITY NODECFG[DEFAULT] PRIORITYF=JOBCOUNT still does the same as it did out of the box. It assigns new nodes rather than filling up existing nodes (see the qsub script below). Thus, let's say I have node1/0, node1/1, node2/0, node2/1, node3/0, node3/1, etc. Under the current configuration, maui just stupidly assigns node1/0, node2/0, node3/0, etc, rather than node1/0, node1/1, node2/0, node2/1, node3/0, node3/1, etc. How should I correctly configure maui to follow this sequential assignment order? Here is an example qsub script. #!/bin/tcsh #PBS -l walltime=1:00:00 #PBS -l mem=400mb #PBS -l nodes=1 # How many nodes did we ask for? set NN = `cat $PBS_NODEFILE | wc -l` lamboot /usr/local/lam-7.1.1-64bit-gnu/bin/mpirun -np 1 /home/wsm/bench/pi lamclean lamhalt -- = Clarence |
|
From: Roger D. <rog...@gl...> - 2005-07-29 08:17:19
|
Hi, I run a build system and I'm thinking of using the mauischeduler as the centerpiece. However, I have three questions: 1. How well do the nodes run on windows boxes? More importantly, do people run the nodes on windows and if so, has anyone any tips on this procedure? 2. I have several linux boxes, some macosX boxes and some windows boxes. Is it possible to give the mauischeduler a job for windows machines only, then get it to sort out which boxes to send the job to? How best can I approximate this? Can I extend the scheduler to cover this? 3. I notice that the pages mention migrating to a more open way of communicating btwn scheduler and nodes. Is there any progress on this? TIA Roger |
|
From: <ga...@ms...> - 2005-05-13 20:54:29
|
On Fri, May 13, 2005 at 02:30:36PM -0600, Daniel A Hansen wrote: > I work an engineering computer lab at BYU university. The lab is > composed mostly of HP-UX computers. We have a pbs-maui setup > (supercluster.org version). The basic idea of our setup is to utilize > the computers in the lab as a cluster when they are not being used by > students. Once a student logs in to use the workstation, we want all of > the jobs running on that mom to be transferred to another mom. This is > where the problem comes in, because if the job is killed then it goes > back into the pbs queue, but maui blocks the job, and so it never gets > assigned to another mom. Any suggestions would be greatly appreciated. > Thanks. > Actually, i should have read your question more thoroughly :\ The cycle harvesting in PBSPro will suspend the jobs when someone logs in, not move them to another node. The jobs will then resume running when the keyboard and mouse have been idle for a number of minutes (we use 10 in our lab). Gabe -- Gabe Turner ga...@ms... UNIX Systems Administrator, University of Minnesota Supercomputing Institute for Digital Simulation and Advanced Computation www.msi.umn.edu |
|
From: <ga...@ms...> - 2005-05-13 20:51:42
|
On Fri, May 13, 2005 at 02:30:36PM -0600, Daniel A Hansen wrote: > I work an engineering computer lab at BYU university. The lab is > composed mostly of HP-UX computers. We have a pbs-maui setup > (supercluster.org version). The basic idea of our setup is to utilize > the computers in the lab as a cluster when they are not being used by > students. Once a student logs in to use the workstation, we want all of > the jobs running on that mom to be transferred to another mom. This is > where the problem comes in, because if the job is killed then it goes > back into the pbs queue, but maui blocks the job, and so it never gets > assigned to another mom. Any suggestions would be greatly appreciated. > Thanks. > This is pretty commonly known as 'cycle harvesting' and is supported in PBS Pro. As far as I know, this function is in neither TORQUE nor OpenPBS. -- Gabe Turner ga...@ms... UNIX Systems Administrator, University of Minnesota Supercomputing Institute for Digital Simulation and Advanced Computation www.msi.umn.edu |
|
From: Daniel A H. <dh...@by...> - 2005-05-13 20:30:37
|
I work an engineering computer lab at BYU university. The lab is composed = mostly of HP-UX computers. We have a pbs-maui setup (supercluster.org ve= rsion). The basic idea of our setup is to utilize the computers in the l= ab as a cluster when they are not being used by students. Once a student= logs in to use the workstation, we want all of the jobs running on that = mom to be transferred to another mom. This is where the problem comes in= , because if the job is killed then it goes back into the pbs queue, but = maui blocks the job, and so it never gets assigned to another mom. Any s= uggestions would be greatly appreciated. Thanks. Daniel Hansen |
|
From: Aaron S. <as...@mh...> - 2004-08-12 00:34:49
|
Hi- This is really a question for the supercluster.org lists. My experience is that should be quite do-able with any variant of the Maui Scheduler. Good luck! On Wed, Aug 11, 2004 at 04:13:04PM -0700, Scott Webster wrote: > > I have a situation with our cluster where we have a limited number of > USB hardware keys required to run a certain piece of software. What > I'd like to be able to do is plug the key(s) into a node or two, flag > that node with some "HAS_KEY" identifier, and then get the scheduler > to put jobs that require that key onto those nodes only. These > special nodes would be used for non-special jobs when there were no > other regular nodes available... > > Is this possible/easy? I'm a bit new to maui, but have poked around a > bit and can't seem to figure it out. I keep seeing references to this > sort of thing on supercluster.org/clusterresources.com but I take it > this is some sort of commercial scheduler that uses maui? > > Any advice or just a point in the right direction would be great! > > Thanks, > > Scott Webster > > PS I think we're running version 1.23 > |
|
From: Scott W. <swe...@ph...> - 2004-08-11 23:13:10
|
I have a situation with our cluster where we have a limited number of USB hardware keys required to run a certain piece of software. What I'd like to be able to do is plug the key(s) into a node or two, flag that node with some "HAS_KEY" identifier, and then get the scheduler to put jobs that require that key onto those nodes only. These special nodes would be used for non-special jobs when there were no other regular nodes available... Is this possible/easy? I'm a bit new to maui, but have poked around a bit and can't seem to figure it out. I keep seeing references to this sort of thing on supercluster.org/clusterresources.com but I take it this is some sort of commercial scheduler that uses maui? Any advice or just a point in the right direction would be great! Thanks, Scott Webster PS I think we're running version 1.23 |
|
From: Aaron C. <as...@mh...> - 2004-05-26 21:25:08
|
Hi Brett- The scheme you envision is do-able but prolly requires a bit of hacking on your own part. We will welcome any patches and/or code contributions to support this endeavor. Regards- Aaron > Hi, > > I'm considering using MauiME on a "comunal cluster" I am building. In > this cluster, different groups contribute different number of compute > nodes with the expectation of a priority based on their level of > contribution. > > Specifically, if a group happens to be the sole user at any particular > time, they have access to all nodes. However, if there are multiple > groups running, each group should have a priority proportional to > their relative contributions (relative to the contributions of just > those groups current running, not of all contributors). > > Is such a scheme possible or is a custom scheduler needed? > > -Brett. |
|
From: Brett V. <bv...@bn...> - 2004-05-25 00:32:07
|
Hi, I'm considering using MauiME on a "comunal cluster" I am building. In this cluster, different groups contribute different number of compute nodes with the expectation of a priority based on their level of contribution. Specifically, if a group happens to be the sole user at any particular time, they have access to all nodes. However, if there are multiple groups running, each group should have a priority proportional to their relative contributions (relative to the contributions of just those groups current running, not of all contributors). Is such a scheme possible or is a custom scheduler needed? -Brett. |
|
From: Aaron S. <as...@mh...> - 2003-12-05 19:36:53
|
On Fri, Dec 05, 2003 at 02:00:04PM +0100, Mag...@as... wrote: > Hi, I'm a newbie in the Maui world who gonna start to use Maui within a > Linux cluster together with SGE. > I want to retrieve information from Maui and noted that there's a nice Java > API for these kinds of things. > However, I also noticed a similar Maui project written in C/C++ and wonder > which Maui code to actually run in > our cluster? Is the functionality the same as well as performance? The Java version (MauiME) is different from the version written in C/C++; this mailing list concerns the Java version. We have run MauiME on our 256 node, 512 processor Linux cluster without any performance problem. We running using the latest Sun Linux JDK 1.4.X MauiME requires at least JDK version 1.2. You may have success running it with the latest Kaffe JVM or with IBM's JVM. |
|
From: <Mag...@as...> - 2003-12-05 13:01:45
|
Hi, I'm a newbie in the Maui world who gonna start to use Maui within a Linux cluster together with SGE. I want to retrieve information from Maui and noted that there's a nice Java API for these kinds of things. However, I also noticed a similar Maui project written in C/C++ and wonder which Maui code to actually run in our cluster? Is the functionality the same as well as performance? Regards, Magnus Bjork |
|
From: Aaron C. <as...@mh...> - 2003-07-01 20:11:38
|
Robert- This is the wrong list to ask this question. You may want to try around at www.supercluster.org. Good Luck! |