Another question on InfluxDB. I have been running the MongoDB/InfluxDb combination for a while now and have noticed the following:
1. I am getting a significant number of CommandResponse events recorded with a single Originating Event being "06e29441-a2f7-4b2c-9f89-ebcdf0" these appear to be in response to the Measurement events sent from the devices. Although, I am not able to find where this is configured, and they are not been generated anywhere by our devices.
2. While these events are not a problem in themselves, the way they are been written into the InfluxDB is causing a significant increase in the number of Series, i.e. increasing the Cardinality of the database, and therefore the memory requirements. The respEvent field, which contains the eid of the originating event, i.e. the measurement event in this case, is essentially unique for each of these Events. As this is written into the InfluxDD as a TAG and not a FIELD, every record is creating a new Series in the InfluxDB. These Series are growing at the same rate as the Events received, already at over 60k in our case with only a few devices logging data for about a week. Unless there is a compelling reason to write the respEvent as a TAG, it would be better to write it as FIELD, in which case there would be a smaller number of Series, with 10s of thousands of data points in each, which is better aligned with InfluxDB best practice.
Any views around whether these CommandResponse Events are standard and/or can be configured in anyway, as well as how to deal with how they are been recorded in the InfluxDB will be appreciated.