Annual Results 2000 - NIKHEF：2000年度业绩- NIKHEF.pptVIP

Pos429.ppt 970234 Remote Online Farms Sander Klous Large Hadron Collider and ATLAS Data processing nightmare There is no way to store all the info produced by ATLAS 40 million events per second x 1.5 MB/event = 60 TB per second In fact: 99.9995% of the data is thrown away So…The data processing nightmare is all about storage Unfortunately… No Rigorous multilevel trigger system First level in hardware Higher levels in software But what if your favoritechannel is not in the 0.0005%? Online bottleneck Well, that’s a problem… So…The data processing nightmare is all about CPU Some are more equal than others… Detector calibration Physics selection Economics In fact, it is about balance Maximize performance ? Minimize costs Is it worth the effort? Gary Stix, editor of Scientific American This can be difficult… The basics Stream implementation Grid/Proxy implementation Open issues Data management Software management Database access Authentication, Authorization and Accounting Performance and reliability Looking for PhD student… If you are interested: Mail to sander@nikhef.nl Conclusion Remote online farms are interesting From a physics perspective From a computer science perspective From an organizational perspective Infrastructure is put in place Many open questions More news next year PhD candidates: mail to sander@nikhef.nl 01 11 010 001 1101 1110 11001 01011 110110 001101 1111111 011100001001110 110111001 000101101 1111010001 0101111100 111101001111 010110000101 H t W Z ? 0 On behalf of the Remote Online Farms Working Group ACAT 2007 25 April LVL1 HLT Scarce CPU resources Networking enables us to prioritize these activities So… The data processingnightmare is all aboutnetworking Amsterdam NIKHEF/SARA Data Acquisition 40 MHz Level 1 Level 2 Accept 1 in 500 Accept 1 in 50 Accept 1 in 10 Level 3 Computing grid Network switch January 2001 ROB Data Collection Network L2PU EF Event Filter North Area ATLAS Detectors Level 1 Trigger Back End Network SFO Mass st