Description:
------------
I'm trying to tune up my application, I make heavy use of the user cache mainly holding stat()s to decide when to rebuild images etc, on rebuild I apc_store the date they were built.  To further optimize my appication I want to know how many calls are misses and whether some kind of cache priming would be good.

Maybe I'm missing something but....



Reproduce code:
---------------
<?
for ($i=1; $i<1000; $i++){
  apc_fetch("I dont exist - I should be a cache miss");
}
?>

User Cache Information
Cached Variables 6868 ( 9.2 MBytes) 
Hits 33136 
Misses 0 
Request Rate (hits, misses) 22.54 cache requests/second 
Hit Rate 22.54 cache requests/second 
Miss Rate 0.00 cache requests/second 
Insert Rate 7.17 cache requests/second 
Cache full count 0 


Expected result:
----------------
apc_fetching something that doesn't exist should be counted as a cache miss?

Actual result:
--------------
My request rate, is always the same as the hit rate (100%), yet 30% of them were apparently inserts, so my code would have got a miss back from apc_fetch and so decided to apc_store the new value.

My user cache miss rate is always zero.