Assume I have 2 CPU's C1 and C2. At the very base, every process executes on the CPU for a given time slice 't'. Using cgroups, assume I limit the process to ensure that it takes only 50% of cpu. Now, when this process runs on C1, once 't' is over, this process is preempted from the cpu and another new process starts executing. Now what happens to the evicted process?
Is it scheduled onto another CPU C2 as soon as it is evicted from C1?
- If it does, will it not get the same effect as allotting 1 CPU?
- If it doesn't how is this execution managed? how long will it take before execution resumes ? Is there any way we can change the configuration for this time period ?