Notes on Singularity

November 14, 2017

  • no singularity module on ifarm1101 (CentOS6)
    ifarm1101:marki:marki> module load singularity
    ModuleCmd_Load.c(204):ERROR:105: Unable to locate a modulefile for 'singularity'
    
  • user-space singularity build did not work:
    > singularity shell /scratch/marki/centos7_farm.img
    ERROR  : Failed invoking the NEWUSER namespace runtime: Invalid argument
    ABORT  : Retval = 255
    

    Seems to need to be installed by root. See https://github.com/singularityware/singularity/issues/415

  • singularity is a standard package under fedora 26, RHEL7

 

 

 

Advertisements

GlueX Meeting Report, November 7, 2017

  • New work disk
    • 60 TB being rsync’ed now
    • started last Thu.? Fri?, 1/3 done
    • deleting files appreciated
  • amptools 0.9.3, sim-recon 2.19.0 released
  • Physics will initiate another procurement of Lustre-based disk, 200 TB worth, to augment our volatile and work space. There is the possibility for more if we can justify it.
  • This afternoon: Computing Round Table, 1 pm, F113,
    • Andrea Dotti (SLAC), Frank Gaede (DESY), and Brett Viren (BNL) will discuss the geometry and detector description from the perspectives of the Geant4, DD4hep, and GeGeDe projects.

 

differing scons behavior under tcsh vs. bash

Note added 10/17: could not reproduce, moving on.

bash:

g++ -pthread -shared build_hddm_mc_s/temp.linux-x86_64-2.7/pyhddm_mc_s.o build_hddm_mc_s/temp.linux-x86_64-2.7/u/group/halld/Software/builds/Linux_RHEL6-x86_64-gcc4.9.2/sim-recon/sim-recon-2.18.0/src/.Linux_RHEL6-x86_64-gcc4.9.2/libraries/HDDM/hddm_mc_s++.o -L/group/halld/Software/builds/Linux_RHEL6-x86_64-gcc4.9.2/sim-recon/sim-recon-2.18.0/Linux_RHEL6-x86_64-gcc4.9.2/lib -L. -lHDDM -lxstream -lbz2 -lz -lpython2.7 -o build_hddm_mc_s/lib.linux-x86_64-2.7/hddm_mc_s.so
/usr/bin/ld: cannot find -lpython2.7
collect2: error: ld returned 1 exit status
error: command 'g++' failed with exit status 1
scons: *** [.Linux_RHEL6-x86_64-gcc4.9.2/libraries/HDDM/hddm_mc_s.so] Error 1
scons: building terminated because of errors.

tcsh

g++ -pthread -shared build_hddm_mc_s/temp.linux-x86_64-2.7/pyhddm_mc_s.o build_hddm_mc_s/temp.linux-x86_64-2.7/u/group/halld/Software/builds/Linux_RHEL6-x86_64-gcc4.9.2/sim-recon/sim-recon-2.18.0/src/.Linux_RHEL6-x86_64-gcc4.9.2/libraries/HDDM/hddm_mc_s++.o -L/group/halld/Software/builds/Linux_RHEL6-x86_64-gcc4.9.2/sim-recon/sim-recon-2.18.0/Linux_RHEL6-x86_64-gcc4.9.2/lib -L/u/group/halld/Software/builds/Linux_RHEL6-x86_64-gcc4.9.2/python/Python-2.7.13/lib -lHDDM -lxstream -lbz2 -lz -lpython2.7 -o build_hddm_mc_s/lib.linux-x86_64-2.7/hddm_mc_s.so

Changes to the cache deletion policy

Date: Wed, 16 Aug 2017 13:28:06 -0400 (EDT)
From: Ying Chen <ychen@jlab.org>
To: marki <marki@jlab.org>, Alexander Austregesilo <aaustreg@jlab.org>,
Sean Dobbs <sdobbs@jlab.org>, pmatt <pmatt@jlab.org>

All,

I have done the implementation of the changes to the cache management policy.
Here are the summary of the changes:

1) All files used in farm jobs (as input starts /mss/…) and bigger than 3 MB
will be deleted first.

2) The pin from farm job will not hold due to pin quota. This means even there is
no enough pin quota, the request will still be processed.

3) When the pin excess the pin quota, most near expired pin will be removed
(not necessary the oldest pin).

4) When there is no pin quota available, user’s pin will fail and user’s jcache
wil be hold (same as like before).

I hope these changes will make most pin unnecessary so you don’t have to pin
a large amount of files. In next few days, I will reduce halld’s pin quota gradually
to test the new software, eventually drop to 150TB.

Thanks for your patient and support.

Ying


Note added on point (3): Files greater that two days from pin expiration will not be unpinned. This means that that the group can be over its pin quota.

Note on point (4): Farm pins will always be granted, event if group is over its pin quota. Farm pins are likely to disappear shortly; they are removed when the associated job is finished.