|
Is it not a measure of process loads, meaning that there is no waiting queue if the load is <1 (all processes are served immediately), while the delays increase with loads above 1?
I have had loads around 40-50 before during peak times, but with some optimization I was able to decrease it to 5-8.
|