pm = dynamic pm.max_children = 700 pm.start_servers = 20 pm.min_spare_servers = 20 pm.max_spare_servers = 40 pm.max_requests = 500
I am using the following. But still i am getting errors after reaching certain level of traffic. The Ram and CPU are free.
max_children * max_request = total hits? But i started to get warning after reaching 75 hits per second.
[pool www] server reached pm.max_children setting (5), consider raising it
What will happen if i change the same setting to
pm=ondemand pm.max_children = 700 pm.max_requests = 500
ondemand is not creating any children at startup and started to serve one by one when needed till it reaches 700.
in pm=dynamic, certain child process will be created and and be in waiting status to handle the traffic. In the end, if the traffic is reached, both would have ended with 700 children
How does these improve the performance?
pm.process_idle_timeout = 10s; the default value
pm=dynamic, the min and max spare servers are also confusing. for 700 children, how to calculate max spare servers? what will the 40 spare servers do here with 700 children?
In simple, does it mean that a children will handle 500 request per second and waits for 10seconds and dies?
Please someone give me a good understandable knowledge