conScaleTest2.java is a connect stress test that makes 2000
connections to taller.eng on port 13. The test hangs at the
command prompt. This was originally seen on the test run on
W2000 machine "isolinear" and duplicated on W2000 machine "belagio".
It was also seen on WNT "duotronic".

The same test executed on w98, solaris, solarisIA, and linux



without hanging.

see attached test.


krystyna.polomski@Eng 2001-06-14
B67 update: test case replaced by a better suggested
version;

Solaris is still successful however now win98 hangs.
here is a thread dump from win98:

K:\mmerlin2\tests\nio>java madbot_cST2
Full thread dump:

"Signal Dispatcher" daemon prio=10 tid=0x006A9220 nid=0xfff96ea9 waiting on moni
tor [0..0]

"Finalizer" daemon prio=9 tid=0x006A64A0 nid=0xfff7058d waiting on monitor [89af
000..89afda0]
        at java.lang.Object.wait(Native Method)
        - waiting on <028B0138> (a java.lang.ref.ReferenceQueue$Lock)
        at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:111)
        - locked <028B0138> (a java.lang.ref.ReferenceQueue$Lock)
        at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:126)
        at java.lang.ref.Finalizer$FinalizerThread.run(Finalizer.java:162)

"Reference Handler" daemon prio=10 tid=0x006A50F0 nid=0xfff64735 waiting on moni
tor [894f000..894fda0]
        at java.lang.Object.wait(Native Method)
        - waiting on <028B0198> (a java.lang.ref.Reference$Lock)
        at java.lang.Object.wait(Object.java:425)
        at java.lang.ref.Reference$ReferenceHandler.run(Reference.java:110)
        - locked <028B0198> (a java.lang.ref.Reference$Lock)

"main" prio=5 tid=0x006A1A20 nid=0xfffdcde9 runnable [57f000..57fab4]
        at sun.nio.ch.PollArrayWrapper.poll0(Native Method)
        at sun.nio.ch.PollArrayWrapper.poll(PollArrayWrapper.java:184)
        at sun.nio.ch.SelectorImpl.doSelect(SelectorImpl.java:97)
        - locked <0282B0D8> (a sun.nio.ch.SelectorImpl$Lock)
        at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:120)
        at madbot_cST2.<init>(madbot_cST2.java:77)
        at madbot_cST2.main(madbot_cST2.java:161)

"VM Thread" prio=5 tid=0x006A4470 nid=0xfff923cd runnable

"VM Periodic Task Thread" prio=10 tid=0x006A81C0 nid=0xfffb56a1 waiting on monit
or
"Suspend Checker Thread" prio=10 tid=0x006A8080 nid=0xfff7185d runnable




krystyna.polomski@Eng 2001-06-15
new behavior problem for Solaris for the updated test:
throws a intermittent hotspot error after a fresh compile.
on 6/14 I logged a hotspot bug (4470310) based on an hour long
stress test based on this updated test. Today I received two
hotspot errors but no log file for the updated test:

first error (screen output):
***********
failed - unexpected exception thrown java.lang.NullPointerException

Segmentation fault
************


second error (screen output):
(I was able to force this a second time after ~ 5 re-compiles;
identical error ID)

#
# HotSpot Virtual Machine Error, Internal Error
# Please report this error at
# http://java.sun.com/cgi-bin/bugreport.cgi
#
# Java VM: Java HotSpot(TM) Client VM (1.4.0-beta_refresh-b68 mixed mode)
#
# Error ID: 53484152454432554E54494D450E43505000F6 01
#
# Problematic Thread: prio=5 tid=0x2b540 nid=0x1 runnable
#

[end]


###@###.### 2001-08-23
There is no Linux hang on build77. We
believe another test caused the omission
of the test result initially.

confirming hang on windows in build77.
still a problem.
/krys


###@###.### 2001-10-11
-restoring original synopsis: ...hang on W2000 and WNT, build65
-verifying as fixed on W2000 and WNT build 82
-see new bug filed for new OS failures in b82 for win98 and
Solaris: 4512632
/krys