Kapacitor batch backfilling

I have understood the possibility to backfill a Kapacitor rule based on a replay.
What I would like to know if there is a possibility to backfill a Kapacitor rule acting as a continuous query and based on Influxdb data, so batch only off course.
In my case I have a ratio I cannot compute directly in InfluxDB. Like Calculate Rates across joined series + Backfill | InfluxData Documentation Archive
Sometimes I will need to update the historical data in InfluxDB and would like to be able to apply it in the historical data of my ratio…

It seems I can create Kapacitor record in the past. I was thinking it was only realtime recording…
Since I can replay a tick script on a record, I think my need is fulfil…
It should still be easier to run it in the past without creating a record but if it works I am happy anyway.

Kapacitor has two components that make this possible:

  1. The replay engine, this allows you to replay old data against a task
  2. The recording engine that allows you to save snapshots of data for reuse.

There is also replay-live engine that really just combines to two and skips the storing of the recording. This essentially gets you what you want.

Example CLUI usage:

kapacitor replay-live batch -task NAME_OF_TASK -past 90d -rec-time

The above command runs the batch task off the historical data of the past 90 days.

1 Like