Show CPU/Memory/Disk usage for jobs

[lucyb]

Having this data in Job Server and also within Grafana will allow us to:
(1) proactively manage user experience and therefore reduce copilot support burden
(2) scale the service to prevent critical resource exhaustion and the resulting firefighting and sadness

We believe that Job Runner can provide this information in the data it sends up to Job Server. Job Server can then store this data with the rest of the information about a Job.

We don't yet know what sort of aggregation to do to provide an accurate measure of CPU/memory/disk, so we should start with something easy to do and build up from there. Once it's stored in the Job Server database it will be easy to show in Grafana.

Note: surfacing this information in Job Server is out of scope for this ticket.

[bloodearnest]

The first part of this is done in opensafely-core/job-runner#691

There's a separate ticket is in opensafely-core/job-server#3867