For some reason jstatd processes launched in
sun/tools/jstat/jstatdServerName.sh test are not always killed.

here is the output of "ps -ef | grep java" on vm-8450r-4.sfbay

> date
Fri Nov 30 08:47:11 PST 2007

> ps -ef |grep java

gridadm 18720 10396 0 Nov 25 con 2d05 C:\local\common\jdk\baseline\windows-i586\bin\jstatd -J-Djava.security.policy=C:/local/common/testbase/jtreg/MISC_REGRESSION/test/sun/tools/jstatd/all.policy -p 2098

gridadm 1804 10396 0 Nov 25 con 2d05 C:\local\common\jdk\baseline\windows-i586\bin\jstatd -J-Djava.security.policy=C:/local/common/testbase/jtreg/MISC_REGRESSION/test/sun/tools/jstatd/all.policy -p 2099 -n SecondJstatdServer

According to jstatdServerName.sh the exit of the test is actually trapped:
    34 cleanup() {
    35 kill_proc ${JSTATD_1_PID}
    36 kill_proc ${JSTATD_2_PID}
    37 }
    38
    39 trap 'cleanup' 0 HUP INT QUIT TERM

So, it is not clear whether the problem is related to trap or there are some problems in
kill_proc() function (sun/jvmstat/testlibrary/utils.sh)