https://issues.apache.org/bugzilla/show_bug.cgi?id=51888
--- Comment #33 from Sebb <[email protected]> 2011-09-27 12:44:40 UTC --- (In reply to comment #32) > It means shutdown of Conn Manager can occur while a sample is running (if > server takes some time to respond). > This will give an error like this I think: > 2011/09/27 13:50:15 INFO - > org.apache.jmeter.protocol.http.sampler.HTTPHC4Impl$3: I/O exception > (java.net.SocketException) caught when connecting to the target host: Software > caused connection abort > 2011/09/27 13:50:15 INFO - > org.apache.jmeter.protocol.http.sampler.HTTPHC4Impl$3: Retrying connect I think that's a bug in HC4 - it should not retry the connect once the request has been aborted, which is what I think is happening here. > > To reproduce put a Breakpoint in > org.apache.http.impl.client.DefaultRequestDirector line 393 just before > stopping test, then let run() method end, add breakpoint line 323 of > JMeterThread and let JMeterThread end then let DefaultRequestDirector go and > you will get Error% > 0. > > So it impacts a little the error % but I think it is acceptable. Yes, I'm pretty sure that the error will go away when the HC4 bug is fixed. It should be permissible to call abort() then shutdown() without generating a connect error message from HC4. The hack I added fixes the problem for request retry; I'm hoping I might be able to extend the fix to connect retries. -- 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]
