preliminary work for https://github.com/sourcegraph/sourcegraph/issues/45745, just to split out of complexity.

this issue should just take care of adding all the boilerplate code required for the new data retention worker.

• the work_handler will enqueue a new record for each series it encounters in a data_retention_jobs table
• the data retention worker will periodically dequeued records from that table and handle it

the handling work should be done in the mentioned above issue

/cc @felixfbecker @vovakulikov