In order to test scaling out of containerized solutions you often need a tiny workaload that consumes CPU without waitios.
This is exectly what this workload is designed for.
It will run for a given duration, or infintely, depending on the parameters and can be used to test scale out scenarios on K8 or serverless container solutions.
CompConsumer being quite trivial, it exposes only the following configuration parameter addressed via environnement variables :
| Env Var Name | Default Value | Purpose |
|---|---|---|
| JOB_DURATION_SECONDS | 60 | Run duration in seconds. Specify -1 for endless processing |
| JOB_MEMORY_MBS | 1024 | Megabytes of memory to allocate on process startup |
The code is even more trivial as the whole implementation is contained in the 30 line python3 script named compconsumer.py.
The Docker image is continously build and stored on the Docker Hub.
In order to run the Greeter, you'll need to install python3 and pip3 then issue the following commands :
export JOB_DURATION_SECONDS=10
python3 ./compconsumer.py