Greg Smith <gsmith(at)gregsmith(dot)com> writes:
> Here's what I get as statistics on the buffer pool after a
> scan when the server is "happy", from a run with 20 clients:
> writes=38.3MB (8.2%) pinned+used=38.3MB (8.2%)
> dirty buffer usage count histogram:
> 0=0.1% 1=0.3% 2=26% 3=17% 4=21% 5+=36%
Interesting, but not real meaningful by itself --- what's the situation
for non-dirty buffers? And how many are there of each?
It might also be interesting to know exactly how many buffers were
pinned at the time the scan passed over them. In theory it should be a
small fraction, but maybe it isn't ...
regards, tom lane