Re: [SQU] Hit rate vs. bandwidth saving

Raise your maximum_object_size to something much larger. We usually use
32MB or 56MB.

This will catch a lot of big downloads, like executables and mp3s, etc.
This will begin to pull BHR closer to DHR.

Just a thought.

Erik Heinz wrote:
>
> Hi,
>
> we are running squid-2.3.STABLE4 on a Linux box with 512 MB RAM and about
> 36 GB disk space which are nearly full (about 95%, although I set cache_dir
> to only 28000, but that's another problem ...).
>
> We have about 250000 requests per day yielding a client data transfer of
> about 1.5 GB per day. What puzzles me is that the average hit rate
> (including all TCP.*HIT) is greater than 50% while the average bandwidth
> saving is only at 10-20%.
>
> A typical scenario for one day:
>
> requests bytes
> -------------------------------------
> TCP_HIT 25959 151.24M
> TCP_REFRESH_HIT 47279 59.89M
> TCP_IMS_HIT 56264 18.29M
> TCP_MISS 107679 1241.85M
> TCP_REFRESH_MISS 6407 60.46M
>
> Is this normal or can it be improved?

                                  --
                     Joe Cooper <joe@swelltech.com>
                 Affordable Web Caching Proxy Appliances
                        http://www.swelltech.com

--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html