Ability to continue collecting tool metrics after an intruption

Customer might experience an interruption to config tool running for a variety of reasons (network jitter, server timeout due to huge number of entities being monitored in combination with under-resourced controller, etc). For long running jobs that might take an hour or so to complete, this is burdensome as customer has to restart again with no guarantee that it will not experience interruption again. They would like to be able to pickup where they left off during the last interruption in case this edge case occurs so to eventually complete the job.

Start by locating the config tool's metrics-collection entry point and reading how a long-running job currently handles interruptions. Define what progress must be persisted and how a restarted job resumes from the last completed work. Done means an interrupted collection can continue without repeating all prior work and eventually complete.

Written by the indexing model from the issue text.