There are a number of long-running tasks whose specifics we need to audit (self-repair tasks being a good example, but also purging releases, snapshots, etc.), but we don't want to a) fill the nexus log with a million line items, or b) lose information at the whim of a customer's log levels. Support frequently gets requests from users asking to understand what happened to specific artifacts.
- Create a service that can accept all these detailed actions, in large volumes
- One log file per run of a given scheduled task
- Output files should be dated, clearly indicate which task was run
- Emit the log file location into the log at the start and end of the scheduled task
- Include task logs in the support tools zip