There is some provisions in Squid already to give the cached copy when
revalidation of the cached content is not possible. This assumes the
content was cached and the client did not force a cache-bypass
(no-cache). It is quite likely this can be improved a bit in the case of
reverse proxies.
In the rproxy branch <http://devel.squid-cache.org/rproxy/> you can find
another approach on how to deal with errors in a reverse proxy. Not
quite what you ask for, but may be interesting anyway.
Regards
Henrik
Brian Hurley wrote:
>
> we have started to look thru how to implement squid for our network
> of hosted sites and I am interested if anyone has created a dynamic
> error handling addition. I am wanting to prevent an error page,
> due to server crash or the like, from being cached and displayed.
> it seems that the best way is for squid to know the new content is
> an error and not to clear the cache to bring this content up as new.
> I have not read anywhere that squid can do this, and figured someone
> out there must have wanted to same functionality and is willing to
> talk about it, etc...
>
> thanks
>
> Brian Hurley
> Director - Software Development and QA
> LMiV
> Taking media further
> http://www.lmiv.com
> 300 Meridian, Suite 1800
> Indianapolis, IN 46202
> 317.238.2429
> brian.hurley@lmiv.com
Received on Tue Jun 25 2002 - 16:41:54 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:15:42 MST