I need to do simulation on a large network running for a long time. In this case the trace file will be too large to store and also difficult to deal with during post-processing. I tried to use queue-monitor, but it only helps give me delay and drop rate analysis on queues, rather than flows. Is there any way I can analysis without a large trace file? Thanks, Tingshan Huang -- View this message in context: http://old.nabble.com/-ns--how-to-avoid-large-trace-file--tp29567024p29567024.html Sent from the ns-users mailing list archive at Nabble.com.