On Wed, 2004-06-02 at 17:58, Mark R. Diggory wrote: > So far I'm looking at something like: > > HttpConnection connection = > client.getHttpConnectionManager().getConnection(client.getHostConfiguration()); > connection.setConnectionTimeout(...) > > But, my big concern is that this is again just bleed over into all the > connections as they are still coming from (and being returned to) a > pool. Is this a correct assumption?
Mark, This is not that much of a problem as HttpClient always resets socket timeout on the active connection prior to executing a method. The problem is socket timeout setting is simply not thread safe. Socket timeout value may change while connection is still being initialized leading to unpredictable results. This is a known limitation of 2.0 API which has already been solved in CVS HEAD. 3.0 API (currently at alpha1) allows socket timeout to be set at the HttpMethod level http://issues.apache.org/bugzilla/show_bug.cgi?id=24154 Oleg > > -Mark > > > ______________________________________________________________________ > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
