JDK-4197549 : rmid on NT hangs when used to test JINI
  • Type: Bug
  • Component: core-libs
  • Sub-Component: java.rmi
  • Affected Version: 1.2.0,1.2.2
  • Priority: P4
  • Status: Closed
  • Resolution: Duplicate
  • OS: windows_nt
  • CPU: x86
  • Submitted: 1998-12-15
  • Updated: 1999-05-11
  • Resolved: 1999-05-11
Related Reports
Duplicate :  
JDK-4211057 - Regression test UseCustomSocketFactory.java Failing
Duplicate :  
JDK-4197666 - Socket descriptors leak into processes spawned by Java programs on windows
Description
At this time this defect is a placeholder: more information to be added later.
What we know now, is that the 224 Jini tests start a single RMID.  On NT 4.0
(and NOT on Solaris) rmid occasionally hangs, and becomes completely unresponsive.  The resulting process is unkillable (even with task manager) and
requires a system reboot.  We will gather more information later on exact
processes, memory usage and so forth.
Comments
EVALUATION I am waiting for more information (e.g. is the failure related to test harness problems?) from Mark to proceed with investigating this problem, I have a few ideas about where to start investigating: This bug might also be related to 4105668 and it also might be related to, expected test harness problems diagnosed by Iris Garcia. lairdd@east 1999-01-06 This is a duplicate of 4197666. I am closing it. ann.wollrath@East 1999-05-11
06-01-1999