From EGIWiki
Jump to: navigation, search
EGI Inspire Main page

Inspire reports menu: Home SA1 weekly Reports SA1 Task QR Reports NGI QR Reports NGI QR User support Reports

Quarterly Report Number NGI Name Partner Name Author
3 NGI_CH Switzerland Andres Aeschlimann



Date Location Title Participants Outcome (Short report & Indico URL)
30.11.10 Bern Swiss Grid Day about 50
13.1.11 Bern AAA Infoday total ~100, Grid ~20


Date Location Title Participants Outcome (Short report & Indico URL)
22.-24.11.10 Prague EMI all hands meeting 3 from NGI_CH Work on Argus
21.-22.1.11 Thessalonikki (GR) COST D37 meeting Sergio Maffioletti (UZH) Discussed possible synergies within comp.chem VO


Publication title Journal / Proceedings title Journal references
Volume number

Pages from - to
Et al?


2.1. Progress Summary

PSI has been able to hire an additional system administrator with main responsibility for the PSI T3. Start of work: Feb 1st 2011

UZH established several contacts within comp.chem VO and within former chemomentum project. Established also contact with relevant comp.chem potential partners in Switzerland

2.2. Main Achievements


  • Participated in staged rollout of gLite CREAM 3.2
  • Reformatted Lustre with ext4 which increased performance of the fs and decreased the load on the servers

UZH advocated their tools (GC3pie also part of the EGI application DB) as a tool to facilitate the migration of existing comp.chem usecases towards the NGI-CH national infrastructure and towards EGI

2.3. Issues and mitigation

Issue Description Mitigation Description
PSI: Some insecurity due to takeover of SUN by Oracle. Our Storage is based on Sun X4500/X4540 and Solaris as dcache pool servers. The newest machines we bought in December, they now have been taken already out of the program by Oracle. Solaris updates seemingly can now only be gotten via a support contract. All of this forces us to reconsider the architecture this year. As queuing system we use SGE where the future with Oracle also is unclear.

Following up development of forked product for SGE. Investigating new storage infrastructure for expansion. Maybe freezing Solaris OS on present architecture and run it as long as we use these machines. Also need to study support costs.

UZH: It is interesting to note that most of the comp.chem group contacted were initially very reluctant on the idea of migrating some of their usecases on grid (being it national infrastructure or EGI)

There is the clear need to assist those groups and collaborate actively with them in order to facilitate the enabling/migration process. These groups seems to be interesting in such infrastructures only for supporting very large throughput; for this, standard available tool are not sufficient.

They need an easy way to control the throughput; for this tools like GANGA, or our G3pie seems to fill this gap. But lot of work has to be done together with these groups in order to migrate their usecases