hi,
> What does a "normal load" mean? Does that mean a time when the queries are
> slow?
yes, we are have slow queries (according to postgresql.log) with such load
> Are you sure the machine really has 48GB of RAM? Because from the vmstat
> output it seems like there's just 32GB.
>
> procs -----------memory---------- ---swap-- -----io---- -system--
> ----cpu----
> r b swpd free buff cache si so bi bo in cs us sy
> id wa
> 0 0 1342168 336936 107636 31353512 0 0 14 183 1911 3426
> 2 1 93 4
>
>
> 1342168 + 336936 + 107636 + 31353512 = 33140252 ~ 31GB
strange.
we paid for 48G :-) and top and free show 48G
/root# free
total used free shared buffers cached
Mem: 49564860 49310444 254416 0 30908 30329576
-/+ buffers/cache: 18949960 30614900
Swap: 20971512 1370960 19600552
> Otherwise I don't see anything wrong in the output. What is the size of
> the database (use pg_database_size to get it)? Did it grow significantly
> recently?
>
there are a number of databases in the cluster on that machine,
in the filesystem it adds up to 271G
cheers gerhard