December 16, 2014 Technical note Open Access

Dal Pra, Stefano

MARC21 XML Export

<?xml version='1.0' encoding='UTF-8'?>
<record xmlns="http://www.loc.gov/MARC21/slim">
  <leader>00000nam##2200000uu#4500</leader>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">user-infn</subfield>
  </datafield>
  <datafield tag="540" ind1=" " ind2=" ">
    <subfield code="u">https://creativecommons.org/licenses/by/4.0/</subfield>
    <subfield code="a">Creative Commons Attribution 4.0</subfield>
  </datafield>
  <datafield tag="100" ind1=" " ind2=" ">
    <subfield code="a">Dal Pra, Stefano</subfield>
    <subfield code="u">INFN-CNAF</subfield>
    <subfield code="0">(orcid)0000-0002-1057-2307</subfield>
  </datafield>
  <datafield tag="041" ind1=" " ind2=" ">
    <subfield code="a">eng</subfield>
  </datafield>
  <datafield tag="245" ind1=" " ind2=" ">
    <subfield code="a">JOB PACKING: OPTIMIZED CONFIGURATION FOR JOB SCHEDULING</subfield>
  </datafield>
  <controlfield tag="005">20231102145658.0</controlfield>
  <controlfield tag="001">143357</controlfield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">publication</subfield>
    <subfield code="b">technicalnote</subfield>
  </datafield>
  <datafield tag="260" ind1=" " ind2=" ">
    <subfield code="c">2014-12-16</subfield>
  </datafield>
  <datafield tag="520" ind1=" " ind2=" ">
    <subfield code="a">&lt;p&gt;The default behaviour of a batch system is to dispatch jobs to nodes having the lower&lt;br&gt;
value of some load index. Whilst this causes jobs to be equally distributed among all the&lt;br&gt;
nodes in the farm, there are cases when different types of behaviour may be desirable, such as&lt;br&gt;
having a completely full node before dispatching jobs to another one, or having similar jobs&lt;br&gt;
dispatched to nodes already running jobs of the same kind. This work defines the packing&lt;br&gt;
concept, different packing policies and useful metrics to evaluate how good the policy is. A&lt;br&gt;
simple farm simulator has been written to evaluate the expected impact on a farm of different&lt;br&gt;
packing policy. The simulator is run against a sequence of real jobs, whose parameters have&lt;br&gt;
been taken from the accounting database of INFN-Tier1. The effectiveness of two packing&lt;br&gt;
policies of interest, namely relaxed and exclusive, are compared. The exclusive policy proves&lt;br&gt;
to be better, at the cost of unused cores in the farm, whose number is estimated. The&lt;br&gt;
possibility of implementing the exclusive policy on a specific batch system, LSF 7.06, is&lt;br&gt;
exploited. Relevant configurations are shown and an overall description of the mechanism is&lt;br&gt;
presented.&lt;/p&gt;</subfield>
  </datafield>
  <datafield tag="773" ind1=" " ind2=" ">
    <subfield code="a">10.15161/oar.it/143356</subfield>
    <subfield code="i">isVersionOf</subfield>
    <subfield code="n">doi</subfield>
  </datafield>
  <datafield tag="542" ind1=" " ind2=" ">
    <subfield code="l">open</subfield>
  </datafield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">899259</subfield>
    <subfield code="u">https://www.openaccessrepository.it/record/143357/files/INFN-14-19-CCR.pdf</subfield>
    <subfield code="z">md5:6bb665129f3afb9c965b54082d9fccfe</subfield>
  </datafield>
  <datafield tag="650" ind1="1" ind2="7">
    <subfield code="a">cc-by</subfield>
    <subfield code="2">opendefinition.org</subfield>
  </datafield>
  <datafield tag="024" ind1=" " ind2=" ">
    <subfield code="a">10.15161/oar.it/143357</subfield>
    <subfield code="2">doi</subfield>
  </datafield>
</record>