On Dec 10, 2007 3:24 PM, Luis Lavena <<a href="mailto:luislavena@gmail.com">luislavena@gmail.com</a>> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="Ih2E3d">On Dec 10, 2007 5:11 PM, Dave Cheney <<a href="mailto:dave@cheney.net">dave@cheney.net</a>> wrote:<br>> Yes - it means using nginx as a proxy to mongrel will actually work !<br>><br>> The current behavior 'stuns' nginx every time mongrel decides it
<br>> doesn't like the request as passed (which is fair, nginx isn't as<br>> strict in its request parsing as mongrel) causing it to try every<br>> mongrel in the balancer group in quick succession. The patch makes
<br>> mongrel behave correctly according to the spec and report the correct<br>> error rather than relying on nginx to pass back a 50x when there are<br>> no mongrel left.<br>><br><br></div>So in theory this patch fix the nginx behavior, but how this affects
<br>other balancers like apache one? (I know is broken, but couldn't miss<br>the importance).<br><br>Based on RFC rules [1], it seems is the correct behavior instead of<br>the unexpected close of the connection.<br><br>
Adding some test to the malformed requests doesn't seem problematic,<br>but I'll love some feedback on cross-balancer results (apache,<br>balancer, delegate, etc).<br><br>[1] <a href="http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html#sec10.4" target="_blank">
http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html#sec10.4</a><br></blockquote><div><br>Good question, Dave, are you able to test other balancers?<br><br> ~Wayne <br></div></div>