Why utilityPoolSize is not defined by maxTasks
I see that regardless on the value of maxTasks
parameter, if utilityPoolSize
is null
than (as it is proclaimed in manual) a number of (N_of_CPUs + 1)
*/python -m everest_agent.start
is running after start of the agent. But why so many "workers"? Suppose that maxTasks
is 1, why we need 9 worker-processes (on 8 CPUs host)?
May be it would be worth to hide this unclear utilityPoolSize
from user and "calculate" it somehow depending on the value of maxTasks
?
Edited by VladimirV