i) This is just to let everyone know that as were are taken data, a significant amount of local tracking is done very close to the runs, The TPC calibrations have also been done for a good set of the data and also fairly recently. These recently reconstructed files can be found in /brahms/data14/data/run04/auau/200/rxxxx/seq.. and on /brahms/data08/data/run04/auau/200/rxxxx/seq. Gtr files as they are being generated will be placed under similar naming scheme. Current scripts for generating these official (preliminary) tracking files can be found in ~/bramreco/run04/auau/200/ltr , gtr ... They are kept in CVS too. There can certainly be a need to change these if so update, store in CVS and inform the list. We strongly request that no other local tracking files are generated and stored on disk under any different naming. We will quickly run out of disk space again if this happens. Single runs for developing code calibrations can of course be placed elsewhere preferentially under /brahms/dataxx/scratch/<name> or /brahms/dataxx/group/<name> Please adhere to such convention so there is a chance to figure out who creates files; no more reduced, reduced03,...etc Please stay off data08..for any other purpose; and data07 as well that is used for the official dAu/pp data (ltr,gtr..) A scheme to keep track of ltr, gtr, dst generated is being developed (hiro,bjoern,fv). For this to work it is important to adhere to naming conventions. ii) LSF queuues, Please be careful how many jobs you start at once. Many jobs that eg. read gtr files and creates trees or dst are i/o limited, so pending on disk system this saturates the band widths. Case in point: today ~73 jobs are running on brahms_crs and the file system handled by rmine003 is inaccesible (see e.g. http://www.rhic.bnl.gov/RCF/UserInfo/Facilities/NFSService/Monitoring/Pong/index.html ) This is really the load as demonstarted as starting and resuming all the jobs. iii) Data14 is full with usefull data. Please do not attempt to write any more ot that file system on behalf of bjoern, hiro, ian and myself Flemming ---------------------------------------------------------------- Flemming Videbaek Physics Department Brookhaven National Laboratory e-mail: videbaek@bnl.gov phone: 631-344-4106 _______________________________________________ Brahms-dev-l mailing list Brahms-dev-l@lists.bnl.gov http://lists.bnl.gov/mailman/listinfo/brahms-dev-lReceived on Sun Feb 22 12:42:46 2004
This archive was generated by hypermail 2.1.8 : Sun Feb 22 2004 - 12:43:08 EST