On Thu, Jul 15, 2010 at 2:01 PM, Jonathan Ellis <jbel...@gmail.com> wrote: > The main problem is not the syscall so much as Java insisting on > zeroing out any buffer you create, which is a big hit to performance > when you're allocating buffers for file i/o on each request instead of > just mmaping things. Re-using those buffers would be possible but > difficult; I think using mlockall to "fix" the mmap approach is more > promising.
Sorry if it is a silly question, but what would be the approach? issue a mlockall with the current set (MLC_CURRENT) before mmap the files? Did someone have an opinion on that http://delivery.acm.org/10.1145/1820000/1814327/p20-kamp.pdf?key1=1814327&key2=0207229721&coll=GUIDE&dl=GUIDE&CFID=97156491&CFTOKEN=66732025 ?. Do you think it could be useful for cassandra to arrange things in the files in order to try to have that kind of 'page afinity' in the data and minimize the page faults trasversing the data? Carlos.
