Application threads were stopped for a long time

Question

Hi, We have got the following problem.
application threads were stopped for a long time after Minor GC hava finished. Minor GC takes usally less than 1sec but a time application threads were stopped tekes often 100 sec.
What's jvm doing then? alternatively we want to get more information when Minor GC occured.
gc.log
33919.643: [GC 33919.643: [ParNew: 376030K->102916K(546176K), 0.1462336 secs] 1953320K->1680371K(4642176K), 0.1465861 secs]
Total time for which application threads were stopped: 0.1557298 seconds
Application time: 1.3021807 seconds
34020.827: [GC 34020.827: [ParNew: 376068K->103166K(546176K), 0.3350141 secs] 1953523K->1681429K(4642176K), 0.3353194 secs]
Total time for which application threads were stopped: 100.0714587 seconds
Application time: 1.0932169 seconds
34121.180: [GC 34121.180: [ParNew: 376318K->100333K(546176K), 0.3740967 secs] 1954581K->1680438K(4642176K), 0.3744228 secs]
Total time for which application threads were stopped: 99.2983213 seconds
Application time: 0.7258378 seconds
34122.304: [GC 34122.305: [ParNew: 373485K->115425K(546176K), 0.9584334 secs] 1953590K->1696193K(4642176K), 0.9587701 secs]
Total time for which application threads were stopped: 0.9823952 seconds
machine spec and jvm option
T2000 Solaris10 Sparc
1cpu 8core 32thread 1GHz, 8GB Memory
jvm 1.4.2_14
-d64 -server
-XX:NewSize=800m -XX:MaxNewSize=800m
-XX:SurvivorRatio=1
-XX:+DisableExplicitGC
-XX:+UseConcMarkSweepGC
-XX:+UseParNewGC
-XX:+CMSParallelRemarkEnabled
-XX:MaxPermSize=256m -XX:PermSize=256m
-XX:+UsePerfData
-verbose:gc
-XX:+PrintGCDetails -XX:+PrintGCTimeStamps
-XX:+PrintTenuringDistribution
-XX:+PrintGCApplicationStoppedTime
-XX:+PrintGCApplicationConcurrentTime
-Xloggc:/opt/local/jj/jboss/server/fujiyama/log/gc.log
-Xmx4800M -Xms4800M
we attempt single thread GC or set 2400M for heap memory
or -XX:UseParallelGC or set the following param
-XX:NewSize=1000m -XX:MaxNewSize=1000m
-XX:SurvivorRatio=6
-XX:TargetSurvivorRatio=80
-XX:MaxTenuringThreshold=20
-Xmx4000M -Xms4000M
but the situation wasn't improved.

Answer 1

Hi Thanigaivel
the above flag reports all the stop the world pauses not only
those caused because of gc. Unfortunately, this flag has
misleading name. Thanks for your info!!
Can you think of any reasons other than GC that cause the stop the world pauses? How can I identify the reason? Are there any VM options to log all the activities that causes the "stop the world" pauses?