Collect logs(and other supplementary data) of runtime after execution completed

In most of cases runtime might contain useful for development/debugging logs.
For example if docker or k8s container crashes with stack trace, we have no easy means to retrieve this log.
Probably we need to poll runtime, using combination of data we have - runtime, platform, job_id and in case of kubernetes - "context", retrieve the log and store either with artifacts or in specially designated storage.
Also likely during polling we can know status of runtime execution, if it is successful or for example infrastructure error happened.

Start by tracing how runtime polling uses the runtime, platform, job_id, and Kubernetes context, then inspect how artifacts or designated storage are handled. Determine the scope for collecting post-execution logs and reporting runtime or infrastructure status; done means the behavior and storage location are defined and covered for Docker and Kubernetes failures.

Written by the indexing model from the issue text.