How to understand the CPU graph on a free Google Cloud server?

Question

System administration

e2-micro instances receive 12.5% ​​of 2 VCPUs and are allowed to run up to 2 full VCPUs for short periods of time.

Do I understand correctly that the screenshot uses 3.65% of the resulting piece?

Answer 1

No not like this.
The left graph is the actual utilization of both CPUs. Those. 3.65% is at the peak both CPUs have created such a load.
On the right, the CPU load graph when credits are exhausted will be the maximum allowed load of 12.5% ​​* 2 = 25%
I.e. instances can work quickly, but not for long -) after the "credit" is exhausted, they begin to troll. After the load is below 12.5% ​​on the processor, these credits will accumulate again, and your machine will again be able to take 100% of the processor in peaks.
Pictures (9:10 stress test at 100% cpu):


CPU load graph: