I had a dinner with a friend of mine from CERN who is in the area, he mentioned an interesting usecase: a user run into a bug in ROOT, and as a result ROOT decided to retry writing the same byte to the same location in a file. It was happening in a tight loop. The file had replication turned on (wide-area). As a result, the user pushed 2 petabytes of data through international networks using 95% of dedicated links before someone realized. Since she was writing to the same offset, the resource mgmt didn't detect it. Pretty amazing how things can go wrong... J. ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the QSERV-L list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=QSERV-L&A=1