Dr. Michael Weller wrote:
> Yep, but it doesn't. Anything I can do here? I *MUST* (sorry for
> yelling) get this cached. I need to reduce the download volume and
> there are plenty of people downloading the same files here.
>
> From the context,cc I claim you think this is a bug, is there
> anything I can do to analyze fix it? I just have no clue about
> all the internal bits checked in the http header, and there seems
> no verbose logging output about the processing there.
You are understanding me correctly; I think there migh be a bug or
feature in how Squid calculates staleness for dateless entities.
It should not be that hard to fix I think. All the logics is there, but
presumabely it gets confused somewhere by not having a date..
The heart of thins is http.c:httpCachableReply(), and this is probably a
good place to start when looking into why the object does not get
cached.
Hmm.. reading the code it even has a comment that such replies are not
cached when received from cache peers... I think you will find it if you
look for it. A possible "fix" is to change the "return 0" to "return 1"
some lines below that. This safeguard probably exists to protect from
misbehaving caches.
/Henrik
-- To unsubscribe, see http://www.squid-cache.org/mailing-lists.htmlReceived on Mon Oct 30 2000 - 14:03:54 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:56:02 MST