cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Miles Elam <mi...@pcextremist.com>
Subject Re: invalid caching problem
Date Tue, 20 May 2003 23:38:17 GMT
Not sure why this would be the case.  Does the same thing happen with 
readers?  They use much of the same mechanism for 304s as dynamic 
pipelines.  If yes, the problem is on the Environment side and less in 
the pipeline code.  If no, the opposite is true.

Just from glances at the code, I'd venture a guess that it's the 
former.  I can't find a case where the 304 code would be anything other 
than a shorter code path than a full content generation.  Of course, 
anything is possible I guess, but then again, I'm not seeing 20 second 
return times for either case.

Is anyone else seeing this?

- Miles Elam


Jeremy Quinn wrote:

> On Sunday, May 18, 2003, at 01:22 PM, Jeremy Quinn wrote:
>
>> Strangely, they were taking in the region of 20secs to return, which 
>> seems a bit strange (considering a fresh re-build of the page takes 
>> less than 1 sec ;)
>
> I have more info about this.
>
> I have TomCat+Cocoon behind Apache 2 using mod_proxy.
>
> When I hit a page that comes from a non-expiring pipeline, that I have 
> seen before:
>
> First Cocoon reports 'processed by Apache Cocoon 2.1m2-dev in 568 
> milliseconds'.
> Then after a long pause (the browser is still waiting) Apache sends a 
> 304 saying it took 20 seconds.
>
> I have this happening in Cocoon Samples as well. On two different 
> browsers.



Mime
View raw message