Fli
10-30-2013, 03:43 PM
Hi, how would you explain what is a "load average" or current server load on linux server?
-----------
I read some articles and it should be like this:
Example, You have one dual core CPU, then your load average should be as nearest to 2.00 as possible (as much CPU used as possible). It is same with RAM, more RAM used = faster processing, execution. Unused ram is a dead ram.
If you have Xeon E3, 4 cores, 8 threads, then it means you have total 8 "cores", and your load should be near 8.00, if its lower, then your server is not fully used.
If your load is above your CPU amount, ie 10.00 on Xeon E3 with 8 cores, then it means server is overloaded and that some files waiting in queue for their CPU time..... so this is wrong, because for example website visitors may experience slower website load time or even timeouts..
In another words:
"important thing to understand is that load average do not represent CPU usage. Rather it represent the load on the CPU by processes waiting for their chance to use the CPU."
-----------
I read some articles and it should be like this:
Example, You have one dual core CPU, then your load average should be as nearest to 2.00 as possible (as much CPU used as possible). It is same with RAM, more RAM used = faster processing, execution. Unused ram is a dead ram.
If you have Xeon E3, 4 cores, 8 threads, then it means you have total 8 "cores", and your load should be near 8.00, if its lower, then your server is not fully used.
If your load is above your CPU amount, ie 10.00 on Xeon E3 with 8 cores, then it means server is overloaded and that some files waiting in queue for their CPU time..... so this is wrong, because for example website visitors may experience slower website load time or even timeouts..
In another words:
"important thing to understand is that load average do not represent CPU usage. Rather it represent the load on the CPU by processes waiting for their chance to use the CPU."