Re: Choose cache peer based on Content-Length HTTP header

From: Henrik Nordstrom <henrik@dont-contact.us>
Date: Thu, 14 Dec 2006 00:48:06 +0100

ons 2006-12-13 klockan 18:34 +1300 skrev Craig Appleton:

> Re implimentation maybe sends a HEAD request via the fast link to
> determine the content-length header then proceeds to choose link for
> GET request based on that.
> Something like that.

Maybe could work, but it's a very dangerous thing to do.

The dangerous part is that there is no guarantee the request do not have
significant sideeffects on the server even if the method is a GET. And
additionally there will be quite significant latency involved before the
GET request can be forwarded as it has to wait for the GET to complete.

Another option is to forward the request as usual, and if when response
headers arrive it's found that it is a download abort and retry on the
other link. This should be considerably less dangerous, and does not add
latency to normal browsing.

Regards
Henrik

Received on Wed Dec 13 2006 - 16:48:11 MST

This archive was generated by hypermail pre-2.1.9 : Sat Dec 30 2006 - 12:00:04 MST