[jboss-user] [JBoss jBPM] - Re: jBPM concurrency

[jorgemoralespou_2]

I have tried PVM (in it`s today state) and results are like the following:
100000 executions - 1 Thread


Total exec time
Total executions
Min. exec. time (ms)
Max. exec. time (ms)
Avg. exec. time (ms)

jBpm Run 13113810000001230,3114
jBpm Run 23071510000001090,3072
jBpm Run 32869610000001640,2870	
PVM Run 152201000000290,0522	
PVM Run 254691000000350,0547	
PVM Run 356711000000980,0567	
100000 executions - 10 Thread * 10000 exec/thread


Total exec time
Total executions
Min. exec. time (ms)
Max. exec. time (ms)
Avg. exec. time (ms)

jBpm Run 1278547100000020992,7855
jBpm Run 217222210000007761,7222	
jBpm Run 324198910000009512,4199	
PVM Run 141865100000013800,4187
PVM Run 239873100000012460,3987	
PVM Run 34827810000006170,4828	
100000 executions - 100 Thread * 1000 exec/thread


Total exec time
Total executions
Min. exec. time (ms)
Max. exec. time (ms)
Avg. exec. time (ms)

jBpm Run 111820441000000124711,8204
jBpm Run 2157539210000001513915,7539	
jBpm Run 3214815010000002639321,4815	
PVM Run 1267664100000029392,6766
PVM Run 217423010000001091,7423	
PVM Run 37270910000005900,7271	


This tests are without process persistance. A process with 20 empty nodes, and an action-handler/behaviour configured.

As anyone can see, jBpm doesn`t scale with concurrent usage, but PVM seems to scale fairly well. Hope this stays the same.

Only wish PVM would be in a more usable state.

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4143994#4143994

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4143994