<br>On Wed, Jul 22, 2009 at 10:25 PM, Oliver Adam <span dir="ltr"><<a href="mailto:oadam@madao.de">oadam@madao.de</a>></span> 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;">
This depends. The ServerIron is going to create a 1-to-1 mapping by default. There is a single backend connection for all incoming connection by default. I assume you are talking about a current ServerIron modeul and not somethine like the ServerIron XL (10 years old or even older). There is a feature called TCP offlad to allow backend connections to stay open - multiple clients might use the same backend connection doing so. The ServerIron is going to create the first connection at the time the first request needs to get forrwarded to a given real server. The client is going to close the connection and the ServerIron keeps the backend connection open but closes the frontend connection.<br>
</blockquote><div><br>I need one further clarification -- say a client sends two requests over a TCP connection, the first with Host header '<a href="http://foo.com">foo.com</a>' and a second request with Host header '<a href="http://bar.com">bar.com</a>'. If the ServerIron has been configured to switch based on Host header, are the two requests switched to the right backends? Does this work when clients have HTTP pipelining enabled?<br>
<br>- nagendra<br> <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
Have a look at TCP offload in the documentation - I guess that is what you need based on your question.<br>
<br>
Oliver<div><div></div><div class="h5"><br>
<br>
At 13:53 21.07.2009, Nagendra Modadugu wrote:<br>
<br>
</div></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div><div></div><div class="h5">
I have a proxy that would like to multiplex requests from multiple clients over the same TCP connection (by relying on HTTP 1.1 keep-alive, but not pipelining requests). In general the requests may be from different browsers, so the user agent strings will be different -- does this situation pose a problem for ServerIron configurations that switch to different backends based on user-agent? Or rather, does ServerIron assume a 1:1 connection mapping between frontend and backend connections?<br>
<br>
Please let me know if I can clarify my question. Thanks,<br>
<br>
- nagendra<br></div></div>
_______________________________________________<br>
foundry-nsp mailing list<br>
<a href="mailto:foundry-nsp@puck.nether.net" target="_blank">foundry-nsp@puck.nether.net</a><br>
<a href="http://puck.nether.net/mailman/listinfo/foundry-nsp" target="_blank">http://puck.nether.net/mailman/listinfo/foundry-nsp</a><br>
</blockquote>
<br>
<br>
</blockquote></div><br>