Welcome to the Large System SIG. Our mandate is to promote sharing of experience in building and operating large systems The SIG was formed at the Jefferson Lab meeting of HEPiX in Oct/Nov 2000. The first results were presented to a CERN meeting in the middle of November that year.
Since then, information has been collected on work in progress at various sites with a report at the Paris meeting of HEPiX in April 2001 and an update at the Catania meeting in April 2002.
In May 2001 we held a cluster builders workshop at Fermilab. The overheads of most of the presentations at that meeting are available at the web site in various formats. A summary was also given at CHEP01 and the published paper and the overheads presented are available as is an article which appeared in the September 2001 CERN Courier and also in the May 2002 edition of the IEEE Cluster Computing Newsletter. The full Proceedings are available at the above conference web site.
All the overheads are available at the
web site. and
the Proceedings are currently in preparation.
Site Surveys
Starting in the spring of 2002, we have conducted occasional surveys
across most of the major HEP sites to
gather information about some of their operational procedures. The results
are tables of
1. operator cover
and where lights-out operation has been tried or is in operation and some of
the alarm and monitoring tools in use;
2. power supply
schemes including UPS cover and describing shutdown and startup procedure
handling.
3.
Videoconferencing services covering both central facilities and support
for videoconferencing on desktops.
4.
Anti-virus tools and procedures in force at various sites.
5. A survey of how different HEP sites select their
preferred PC systems.
6. A survey of User
Support facilities at various HEP sites and conclusions.
7. A survey of policies at the different sites for
blocking NETBIOS ports.
8. A survey of
LSF usage at the different sites.
Last update: 14 May 2004