Kapacitor: stream vs. batch

I have a scenario where I would like processing to happen for every new value inserted.

However, there are two constraints:

  • in some scenarios, some data can arrive late. It’s rare, but it does happen.
  • the calculations are time series, so from the time stamp of the last data, until ‘now’ all calculations need to be redone.

Usually data arrives sequentially so it’s not a problem but if some data arrives late, I need to redo calculation from that point until the newest data.

Can I know the latest data / timestamp that was inserted when my script is called in stream mode?