Article Source: Linux average load average meaning
The meaning of load average
The average payload (load average) refers to the average utilization of the operating queue of the system, or it can be considered as the average of the operational processes .
In the case of road conditions, single-core CPUs and single-lane situations are as follows:
- The numbers between the 0.00-1.00 indicate that the traffic is very good at this time, without congestion, and the vehicle can pass without hindrance.
- 1.00 means the road is normal, but it is likely to worsen and cause congestion. At this point the system has no redundant resources, the administrator needs to be optimized.
- 1.00-*** said the road is not very good, if the arrival of 2.00 means there is a bridge on the number of vehicles a few times the vehicle is waiting. You have to check the situation.
Multi-core CPUs-multi-lane scenarios are as follows:
Multi-core CPU, the number of full-load status is "1.00 * CPU Cores", that is, dual-core CPU is 2.00, quad-core CPU is 4.00.
The average process consumes resources such as CPU, memory, disk I/O, and network I/O, in which case the mean load is not a separate CPU usage. That is, memory, disk, network and other factors can also affect the average load value of the system.
In a single-core processor, when the average load value is 1 or less than 1, the system processing process can be very easy, that is, the load is low. When it reaches 3, it will appear very busy, reaching 5 or 8 is not a good process (of which 5 and 8 are currently a controversial threshold, for the sake of conservatism, it is recommended to choose low).
Reference:
Https://www.slyar.com/blog/linux-load-average-three-numbers.html
Http://www.yunweipai.com/archives/2750.html
View load Average data
Load average can be seen in the next few commands
Uptimew
As follows:
of the top command
of the uptime command.
of the W command
The three values of the load average here refer to the system's average load value for the last 1/5/15 minutes, respectively.
According to experience: we should focus on the average load of 5/15 minutes, because the average load of 1 minutes is too frequent, and the high concurrency of an instant can cause a significant change in this value.
For related algorithms please refer to:
Http://itlab.idcquan.com/linux/attestation/816287.html
Linux average load average meaning "turn"