I am increasingly working with InfluxDB as our backend of choice for monitoring data. It is optimized for timeseries data, is schema-less, from my feeling it seems fast (I haven't done benchmarking myself yet), it allows for high compression and provides nice functionality such as (continuous) aggregates with an familiar SQL like query language. (Details: https://docs.influxdata.com/influxdb/v1.4/)
Python has libraries to work with InfluxDB, but for Knime the only interface I can think of is using the HTTP endpoint of InfluxDB. Has anyone used InfluxDB with Knime? It would be nice to have the database nodes integrate InfluxDB but I am not sure of the restrictions (e.g. no joins in InfluxDB). What is your experience?