what makes you think that this will avoid MSS/MTU issues on caches that
don't enforce MSS correctly?
i think it'll have the exact same problem ...
cheers,
lincoln.
At 01:20 AM 30/03/2001 +0800, Miguel A.L. Paraz wrote:
>Before I try this out, I'd appreciate some feedback.
>
>Because of problems with WCCP interception caching, such as MSS/MTU, I'm
>thinking of making end-users use a proxy setting instead, but this proxy
>setting will try to be fail-safe.
>
>For example, customers at a POP will use IP 192.168.0.1 as their proxy in
>their browser. They are free to remove the proxy setting and browse directly
>if they wish. The real 192.168.0.1 is near the core network and will be a
>cluster of caches.
>
>However, to save POP-to-core bandwidth, let's use WCCP to redirect 192.168.0.1
>to a proxy inside the POP. This way, we localize the traffic yet we don't
>surprise the customer if there is proxy trouble since they know there is one.
>
>Thoughts?
>
>
>--
>
>http://www.internet.org.ph The Philippine Internet Resource
>Mobile Voice/Messaging: +63-917-810-9728
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:34 EDT