Root Benchmarks

I am interested in doing some benchmarking to see how fast the structure ttree works for reading, writing and pruning the ttree structures.

Do you know of any studies of this ttree structure from a performance point of view you can point me to? If not, I will write the benchmarking programs myself.

Thanks in advance.

Jon

We have several benchmarks with Trees in $ROOTSYS/test
cd $ROOTSYS/test
make clean
make

Event 1000 0 0 1 (write in no split mode, no compression
Event 1000 0 0 20 (read file )
Event 1000 1 99 1 (write in split mode and compression)
(see comments in MainEvent.cxx, Event.cxx)

stress -b 30
stress -b 1000
bench

Rene