On Feb 23, 2011, at 10:35 AM, Eric Covener wrote: > On Wed, Feb 23, 2011 at 10:12 AM, Gregory Boyce <[email protected]> > wrote: >> Folks, >> Last week I filed this bug report on mod_proxy and I was hoping to get a >> confirmation from you if you agree that this is in fact a bug or if there >> may be something that I'm missing. >> Essentially what I discovered is that when Apache/mod_proxy is acting as a >> reverse proxy in front of a website, and the remote webserver sends a FIN in >> order to close the connection, Apache does not send the corresponding FIN >> packet to fully close the connection until the next time that child process >> is used. > > > The manual could certainly do a better job of describing how the > connection pool is used, with respect to frontend connections (is this > a 2.0 thing only?), child processes, exactly when smax/ttl is checked, > etc. > > Surprising that you managed to burn through all your local ports but > still not managed to trigger that backend connection closure being > noticed -- maybe would make sense with prefork if the pools were > per-process? > > You could also set MaxRequestsPerChild 100k for relief if this is > still a problem. >
couldn't one also use lower level tcp stack tuning to address this?
