https://issues.apache.org/bugzilla/show_bug.cgi?id=45618
Hao Zhong <[EMAIL PROTECTED]> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |REOPENED
Resolution|INVALID |
--- Comment #4 from Hao Zhong <[EMAIL PROTECTED]> 2008-08-14 18:48:32 PST ---
Please take a look at a confirmed bug in JDK 1.5 [1]. This bug is caused by not
explicitly closing a selector. I copy some descriptions from the bug report as
follows.
"However, when SocketAdaptor is done with the temporarily created selector, it
never explicitly calls selector.close(), which causes the Pipe-related sockets
to be orphaned and left in an ESTABLISHED state until process death. Over
time, this exhausts the number of available sockets and degrades both
application and operating system performance. Pipe-related sockets are only
cleaned up when the owning process is killed."
I agree that automatic tools can produce some false positives. As you
suggested, I do not report all the violations but report those violations that
make sense in my opinion.
I am sorry that my tool cannot provide in what situation a bug will cause
problem as my tool relies on static analysis. Still, I believe that you should
take the report into careful considerations because as you can see, not closing
a selector has already cause some problems in JDK 1.5 and they do take efforts
to fix the problems.
[1] http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=5083450
--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
