Re: How is random_page_cost=4 ok?

Greg Smith wrote:

> The drives themselves, and possibly the OS and disk controller, are all 
> running read-ahead algorithms to accelerate this case.  In fact, this 
> *exact* case for the Linux read-ahead stuff that just went mainline 
> recently: http://kerneltrap.org/node/6642

Apparently only the "simple" stuff hit mainline, see [1] and [2], not 
knowing how this turns out for pg-style loads, especially compared to 
the full-fledged patch.

Readahead is probably too much of a beast that no one dares to touch 
with a 3-foot-pole, unless given a large team with good standing in the 
kernel community and concerted regression testing in whatever 
environment Linux is used these days...


michael


[1] http://lwn.net/Articles/235164/
[2] 
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=history;f=mm/readahead.c