MadGraph5_aMC@NLO

Generation of events from gridpack stucks

Asked by Evgeny Soldatov

Dear experts,

I'm trying to run generation from the gridpack for process:
generate p p > vl vl~ a j j QCD=0 QED=5

Gridpack was created without problems.
However the execution of generation from gridpack continues already more than 7 hours...

The current output is:
[esoldato@lxplus032]/tmp/esoldato% ./run.sh 10000 37
Now generating 10000 events with random seed 37 and granularity 1
WRITE GRIDCARD /tmp/esoldato/madevent
P1_qq_vlvlaqq

If I run generation without gridpack stage, it needs just 15 minutes for full run.

I'm adding some process state outputs from the machine:
top
  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
 3877 esoldato 30 10 46064 18m 4400 R 99.9 0.1 21:55.86 madevent

[esoldato@lxplus032]~% ps a | grep esoldato
 3877 pts/73 RN+ 20:10 /tmp/esoldato/madevent/SubProcesses/P1_qq_vlvlaqq/madevent
26459 pts/73 S+ 0:00 /bin/bash /tmp/esoldato/madevent/SubProcesses/P1_qq_vlvlaqq/ajob1

[esoldato@lxplus032]~% ps -ef | grep esoldato
esoldato 3877 26459 99 20:43 pts/73 00:21:09 /tmp/esoldato/madevent/SubProcesses/P1_qq_vlvlaqq/madevent
esoldato 3878 26459 0 20:43 pts/73 00:00:00 tee -a log.txt
esoldato 8313 26523 0 15:16 pts/73 00:00:00 /bin/bash ./run.sh 10000 37
esoldato 8316 8313 0 15:16 pts/73 00:00:00 python ./madevent/bin/gridrun 10000 37 1
esoldato 8388 8316 0 15:16 pts/73 00:00:01 /cvmfs/atlas.cern.ch/repo/ATLASLocalRootBase/x86_64/python/2.7.3-x86_64-slc6-gcc47/sw/lcg/external/Python/2.7.3/x86_64-slc6-gcc47-opt/bin/python -O ./madevent/bin/gridrun 10000 37 1
root 23823 16479 0 21:02 ? 00:00:00 sshd: esoldato [priv]
root 23923 16479 0 21:03 ? 00:00:00 sshd: esoldato [priv]
esoldato 24116 23823 0 21:03 ? 00:00:00 sshd: esoldato@pts/5
esoldato 24140 24116 0 21:03 pts/5 00:00:00 -zsh
esoldato 24198 23923 0 21:03 ? 00:00:00 sshd: esoldato@notty
esoldato 24201 24198 0 21:03 ? 00:00:00 /usr/libexec/openssh/sftp-server
esoldato 25877 24140 0 21:05 pts/5 00:00:00 ps -ef
esoldato 25878 24140 0 21:05 pts/5 00:00:00 grep esoldato
root 26247 16479 0 15:14 ? 00:00:00 sshd: esoldato [priv]
esoldato 26459 8388 0 19:32 pts/73 00:00:00 /bin/bash /tmp/esoldato/madevent/SubProcesses/P1_qq_vlvlaqq/ajob1
esoldato 26497 26247 0 15:14 ? 00:00:00 sshd: esoldato@pts/73
esoldato 26523 26497 0 15:14 pts/73 00:00:01 -zsh

Gridpack can be found here: https://drive.google.com/file/d/1wwRSqqnMavKINzrZxEkMG0zWwuMbo7O1/view?usp=sharing

Do you have any ideas?

With the best regards,
Evgeny

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Which version of MG5aMC?
CMS reported that some gridpack were quite slow since 2.6.1 (due to a change that they request actually)
This is now fixed and going to be released in 2.6.5

Cheers,

Olivier

Revision history for this message
Evgeny Soldatov (esoldato) said : 		#2

Hi Olivier,

I'm using MG5_aMC 2.6.4
The execution is still running (~16 hours already).
I've also tried to run it on other machine (not lxplus) - the same situation...

So the only solution is to wait v.2.6.5, right?
When approximately do you plan to release v.2.6.5?

With the best regards,
Evgeny

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#3

Hi,

I have made the official request to be allowed to release 2.6.5:
https://code.launchpad.net/~mg5core1/mg5amcnlo/2.6.5/+merge/362241
no one object so far so I can do it when I have time to do it (i.e. as soon as I have finished all the validation procedure).
Now my working time on the madgraph project is officially 0% this week. So this will likely be around the end of next week.

Now each build of each version have an associated tar-ball that you can download and use:
so if you go to this page:
https://bazaar.launchpad.net/~mg5core1/mg5amcnlo/2.6.5/revision/304
you will find a tarball with the latest candidate for 2.6.5 release.

Cheers,

Olivier

Revision history for this message
Priscilla Pani (ppani) said : 		#4

Hello,
I'd like to report that I have experienced the same problem (hanging gridpack
event generation) with W with >= 3jet generation and the latest MG version.

I see this problem with 3.0.1 but not with 2.6.5. I attach gridpacks generated in both
versions since maybe this helps reproduce/identify the problem:
2.6.5: https://cernbox.cern.ch/index.php/s/0zzFCehgHehHoIG
3.0.1: https://cernbox.cern.ch/index.php/s/XUvWwcDsXhmwKsp
Please let me know if I can provide more informations. For the moment I
reverted back to 2.6.5 for my jobs.

Cheers and thanks,

Priscilla

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#5

Hi Priscilla,

3.0.1 is based on 2.6.4 and therefore this is expected.
(3.0.1 is a beta version and should not be used for production but thanks for the tests with that version which are obviously useful for us)

Cheers,

Olivier

Revision history for this message
Priscilla Pani (ppani) said : 		#6

Ok perfect, thanks a lot for confirming that this is know .. I will stick with 2.6.5 for production then!

Cheers,

Priscilla

Can you help with this problem?

Provide an answer of your own, or ask Evgeny Soldatov for more information if necessary.

To post a message you must log in.