On 26.09.2011, at 20:42, Stefan Guggisberg wrote: >> However, the overall performance is still a bit low (2:24-3:05 minutes in a >> clean repository). Any idea how the performance could be improved? Am I >> doing something conceptually wrong? > > did you run my test with the same test data (local svn export of jackrabbit > trunk)? >
On svn export of jackrabbit trunk it takes ~1 minute (note that in your code you increased the counter two times in the loop. Thus, I've just half of the units than you had). However, it's factor two slower than on your machine. The size of the attached data, however, does not seem to be the most limiting factor, right? (27 MB & 3k files -> 1 minute; 240 MB & 6k files -> 3 minutes) Mailing List approach: 0:00:06.440: 500 units persisted. data 3 MB 0:00:15.066: 1000 units persisted. data 7 MB 0:00:47.602: 1500 units persisted. data 15 MB 0:00:50.955: 2000 units persisted. data 18 MB 0:00:54.042: 2500 units persisted. data 22 MB 0:01:00.552: 3000 units persisted. data 27 MB Run took 0:01:07.887 Jackrabbit First Hops adapted: 0:00:03.192: 500 units persisted. data 3 MB 0:00:08.062: 1000 units persisted. data 8 MB 0:00:38.184: 1500 units persisted. data 19 MB 0:00:40.160: 2000 units persisted. data 22 MB 0:00:42.568: 2500 units persisted. data 25 MB 0:00:46.677: 3000 units persisted. data 32 MB Mon Sep 26 21:02:15 CEST 2011: 3245 units persisted Run took 0:00:51.384
