Re: memory-mapped files in Squid

On 29 Jan 99, at 12:36, Henrik Nordstrom <hno@hem.passagen.se> wrote:

> Thanks Oskar.
>
> Your TCP_REFRESH_HIT ratio is 67% of 0.0070%.
>
> Of those few refreshes that you had 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).

 thats not that simple. refreshes happen only when the object is requested
 another time. refresh_hit is almost as pure win as plain tcp_hit. Squid
 just decided to go and check with the source. This can only speak about
 the fact that this object has been in cache for some time and/or refresh_rules
 required squid to refresh-check.

 This rule applies also to objects that are requested 1000 times a day,
 so that they remain on the top-10 of the LRU, but its just their age
 that requires squid to do refreshes after some time.

 some very old objects that are reused every 9 weeks can on the other
 hand give plain tcp_hit, and so are not covered with these stats.
 Very many gifs can be fresh for ages, and there's no need to refresh
 them before refresh_max time psaaes

 And, if the whole cache is flooded in a day-or-two, there's almost no
 good reason for these refreshes to appear.

 ----------------------------------------------------------------------
  Andres Kroonmaa mail: andre@online.ee
  Network Manager
  Organization: MicroLink Online Tel: 6308 909
  Tallinn, Sakala 19 Pho: +372 6308 909
  Estonia, EE0001 http://www.online.ee Fax: +372 6308 901
 ----------------------------------------------------------------------
Received on Tue Jul 29 2003 - 13:15:56 MDT