出版社：SISSA, Scuola Internazionale Superiore di Studi Avanzati
摘要：The SFXC software correlator developed at JIVE has reached state of maturity where we are
concentrating on optimization and testing on “real-world” data sets. In this context, distribution
of sampled data is an important issue. We will discuss an approach that makes use of existing
Grid middleware suitable for near-realtime correlation that can potentially be distributed over
multiple clusters. We will also discuss an entirely different approach, more suitable for real-time
correlation where data is streamed directly into a cluster using the GEANT2 AutoBAHN and
Internet2 DCN infrastructure.
