Currently I am evaluating InfluxDB, and just have a question for the query performance: is the query performance linear to the number of query fields? Is there a way to avoid this?
The thing is usually we will have a few thousands sensors to collect time series data. For my current project, I need to query a lot of sensors for each query. To test this, I build a simple measurement with about 6k fields, each field has 86k points. I am using a script to generate a long SQL to query 2000 fields, which takes 90sec on my mac. Is there a better way to do this?