I am using Kapacitor to transform already collected data. The data comes into Tick scripts as a stream, and then I perform some built-in functions and use a UDF to adjust both fields, tags, and the value of time
.
The points coming out of the UDF are in nanoseconds but really only contain actual precision of milliseconds (i.e., every time
values has 6 zeros on the end).
I use InfluxDBOut
to write the new points into a new Database in InfluxDB. Everything works fine, but I was curious about potentially optimizing things by actually setting the precision for the calls in InfluxDBOut
. I tried using the property InfluxDBOut.precision('ms')
but that did not seem to have an effect in the write calls to InfluxDB (i.e., I can see in Kapacitor.log that the write calls to the InfluxDB http API still use precision=ns
).
Is there something about how the time
field is stored in Kapacitor that nullifies my use of InfluxDBOut.precision('ms')
without using stats.align()
or window.align()
? Is there anything I can do in my UDF to help force Kapacitor to use the assigned precision in InfluxDBOut
?