I have a job that polls a webservice every 10 min and basically executes

process_webservice_events(from=previous_timestamp, to=current_timstamp)

The result is idempotent, so calling the function multiple time with the same time window does not alter the result. This is a custom job with custom retry logic and it stores the last timestamp locally. I migrated this almost 1:1 to dagster and I was wondering if you could point out how to improve this with built in functionality from dagster. I read about partitioning, but am not entirely sure this is an appropriate use case.

If you have incremental timestamps, might not be the best use case for partitions (partitions tend to be definable increments without overlaps)

Thanks!