Re: memory-mapped files in Squid

Hi

> Your TCP_REFRESH_HIT ratio is 67% of 0.0070%.
   
> Of those few refreshes that you have there was fewer hits than I
> expected, but still enought to suggest that I am thinking in the right
> direction. However if the refresh ratio commonly is that small there is
> no apparent need to bother with saving disk space on refreshes (less
> than 0.5% estimated space saving).

If I understand this correctly, the major advantage of a fifo buffer is
that there is almost no fragmentation... right? This would mean that the
0.5% waste would be insignificant to the +- 10% of disk space I have to
keep free on my disks at the moment to avoid fragmentation.

I am not sure of the fifo buffer idea, though. I don't believe that any of
the commercial systems are using this... and they have lots of money to
throw at getting a filesystem right.

I personally think that a fairly classic filesystem (something like the sfs
code) is probably the way to go.

Oskar

---
"Haven't slept at all. I don't see why people insist on sleeping. You feel
so much better if you don't. And how can anyone want to lose a minute -
a single minute of being alive?"				-- Think Twice