Allow to specify a cpu pressure usage value

[taihuynh167]

What does this PR do?

• Adds new functionality
• Alters existing functionality
• Fixes a bug
• Improves documentation or testing

Please briefly describe your changes as well as the motivation behind them:

• CPU pressure disruption currently does not take any parameter, and always uses as much CPU as possible.
• This PR enables users to specify a specific value, which will only fill a certain amount of the CPU.
• This amount would be specified as a percentage (of CPU to fill) or as an integer (number of cores to fill).

Code Quality Checklist

• The documentation is up to date.
  • To be updated
• My code is sufficiently commented and passes continuous integration checks.
• I have signed my commit (see Contributing Docs).

Testing

• I leveraged continuous integration testing
  • by depending on existing unit tests or end-to-end tests.
  • by adding new unit tests or end-to-end tests.
    • I plan to separately unit test the StresserManager. The cpu_pressure.go now delegates most of the stuff to StresserManager.
• I manually tested the following steps:
  • locally.

{"level":"info","ts":1666618997157.501,"caller":"cgroup/cgroup.go:59","message":"reading from cgroup file","disruptionName":"cpu-pressure","disruptionNamespace":"chaos-demo","targetName":"demo-curl-79bf7b74c9-4m54c","targetNodeName":"colima","path":"/mnt/cgroup/cpuset//kubepods/burstable/pod31634024-44d6-4da8-b549-b4c8cb624f59/529ff60ee9eec61fe3d5edaca93c6ac8b84c3dc08ef214666da62dc59e77ae51/cpuset.cpus"}
{"level":"info","ts":1666618997157.7095,"caller":"stress/cpuset_manager.go:87","message":"retrieved coresToBeStressed after applying userRequestCount 50%: 0-1","disruptionName":"cpu-pressure","disruptionNamespace":"chaos-demo","targetName":"demo-curl-79bf7b74c9-4m54c","targetNodeName":"colima"}
{"level":"info","ts":1666618997158.0562,"caller":"stress/cpuset_manager.go:49","message":"changed GOMAXPROCS value from 4 to 2","disruptionName":"cpu-pressure","disruptionNamespace":"chaos-demo","targetName":"demo-curl-79bf7b74c9-4m54c","targetNodeName":"colima"}
{"level":"info","ts":1666618997158.147,"caller":"injector/cpu_pressure.go:124","message":"joining target CPU cgroup","disruptionName":"cpu-pressure","disruptionNamespace":"chaos-demo","targetName":"demo-curl-79bf7b74c9-4m54c","targetNodeName":"colima","core":3,"pid":31299}

- [ ] as a canary deployment to a cluster.

[luphaz]

The implementation suggest for every container we'll stress half the assigned CPU

It seems OK from my POV, I would still appreciate more reviews from people having broadest knowledge here

Maybe updating the doc:

• https://github.com/DataDog/chaos-controller/blob/main/docs/cpu_pressure.md

Or FAQ to explain the behaviour could help users of this feature

[aymericDD]

LGTM 👍