Re: maximum_object_size

From: Jon Kay <jkay@dont-contact.us>
Date: Sun, 16 Dec 2001 11:10:35 -0600

Joe Cooper wrote:
>
> Jon Kay wrote:
>
> > I would like to propose the elimination of maximum_object_size, in
> > squid.conf. What does it do, you ask? Well, it sets a maximum
> > object size that can be stored on disk. This seems like
> > completely redundant functionality given the existence of the four
> > [request|reply]_[header|body]_max_size variables.
> >
> > Worse, bringing in a file that exceeds the size of this variable
> > does not result in an error page or log entry. The only thing that
> > happens is that a variable is a silently incremented, making
> > it a very difficult problem to debug. I found out about this
> > variable it by gdb, which seems wrong.
>
> What is the problem that is difficult to debug? Does it cause a problem
> to download something greater than maximum_object_size, or is it just
> that you'd like to know when something isn't cacheable due to size?
>
> Though it does appear to be redundant with the other options you've
> mentioned...

There is nothing wrong with having a maximum storable object size control
per se. The problem is that it is very difficult behavior for users to
even notice, much less understand. That's why it's better to have that
kind of size control be an up-front limitation. Which is why we already
have
a zillion up-front size limit parameters. If we keep it, it should at
least
default to no limitation. But I think it'll just confuse people. Better
to
take it out and let people use the easier-to-understand
[request|reply]_*_max_size
variables.

-- 
Jon Kay        pushcache.com                      jkay@pushcache.com
http://www.pushcache.com/                             (512) 420-9025
Squid consulting				  'push done right.'
Received on Sun Dec 16 2001 - 10:11:26 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:40 MST