The trick is to use vmstat and look at the number for “wait on acknowledge”.

vmstat -S m 1 100

Look under “cpu” and the “wa” column. If this number is high, it is bad. It should be zero. Some sample output from vmstat looks like this:

procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa
 0  0      0    406     83   3062    0    0     1     5    1    4  0  0 99  1
 0  0      0    406     83   3062    0    0     0     0 2020   49  0  0 100  0
 0  0      0    406     83   3062    0    0     0     0 2020   41  0  0 100  0
 0  0      0    406     83   3062    0    0     0     0 2018   53  0  0 100  0

Look at the column furthest to the right.