[
https://issues.apache.org/jira/browse/HTTPCORE-85?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12507637
]
Anders Wallgren commented on HTTPCORE-85:
-----------------------------------------
On further inspection, this looks like it was self-inflicted. Much earlier, my
code failed an assertion, which caused one of the worker threads in the
IOReactor to die. Later on, my guess is, another request was assigned to that
thred, which is why the session request never gets processed.
This does raise the issue, however, of how the IOReactor should behave when
client code misbehaves...I'm of two minds as to whether the need for
robustness, on the one hand, should trump the desire not to mask ill-behaved
code, on the other.
> Some SessionRequests to DefaultConnectingIOReactor get dropped during load
> --------------------------------------------------------------------------
>
> Key: HTTPCORE-85
> URL: https://issues.apache.org/jira/browse/HTTPCORE-85
> Project: HttpComponents Core
> Issue Type: Bug
> Components: HttpCore NIO
> Affects Versions: 4.0-alpha5
> Environment: RHEL 4, dual-core
> Reporter: Anders Wallgren
>
> Some of our tests are still failing, against the current SVN head (Revision
> 549923).
> One of the test that fails intermittently issues two session requests from a
> single thread to a DefaultConnectingIOReactor, one after the other, to the
> same host/port.
> The symptom of the failure is that the first SessionRequest never proceeds,
> while the second does just fine. It happens about 10% of the time.
> I will attempt to pare this down to simple test case.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
