file side memcached

[rm_memcached at cheapcomplexdevices.com]

Why not just run memcached with a larger memory size than physical memory 
and let the kernel's swaping logic handle which blocks and when get 
swapped out?

I'd guess that memcached's memory access pattern (slabs and hash 
tables) are pretty friendly to the virtual memory system, at least 
compared to trees or other structures that need walking.

Currently I note that some memory on the systems where I have memcached 
running is swapped out - not sure if that's memcached memory or not; but 
I'd guess so since the cache is so large at least parts of it are likely 
to be relatively infrequently used.