Open glassfishrobot opened 15 years ago
Reported by kevinconaway
kevinconaway said: Created an attachment (id=16) Attaching patch that uses and ArrayBlockingQueue instead of synchronized access to a Stack
kevinconaway said: Created an attachment (id=17) Attaching updated patch that returns parsers to the pool appropriately in case of error
@vbkumarjayanti said: Thanks for this. Will review your patch and try to put it in ASAP.
@vbkumarjayanti said: Hi kevinconaway,
Sorry i did not act for long on your patch, just got busy with my other project(s).
Now it looks like the idea of parserpools itself is problematic.
Please see : https://saaj.dev.java.net/issues/show_bug.cgi?id=46
I will look into both the issues together and see what is best.
Thanks again for providing the patch. If you have an opinion let me know.
Thanks.
@vbkumarjayanti said: I have putback your proposed changes. Thanks for the same, though i now still need to investigate issue 46.
Thanks.
kevinconaway said: Created an attachment (id=18) Updated patch to set the interrupt status
kevinconaway said: I'm reopening the issue because I've added an updated patch. The ParserPool needs to call Thread.currentThread().interrupt() when catching an InterruptedException. This ensures that the interrupt gets propagated up the call chain.
File: saaj-41-updated-2.patch Attached By: kevinconaway
File: saaj-41-updated.patch Attached By: kevinconaway
File: saaj-41.patch Attached By: kevinconaway
Was assigned to saaj-issues
This issue was imported from java.net JIRA SAAJ-41
The com.sun.xml.messaging.saaj.util.ParserPool does not perform well under heavy concurrent use. This class should be refactored to use the facilities provided by java.util.concurrent.
Environment
Operating System: All Platform: All
Affected Versions
[1.3.1]