With InfluxDB 1.X we have several Kapacitors subscribed to different InfluxDB and installed/running on different hosts, which allow us to select different subscriptions and control the performance and available resources on each service/hosts.
As I have understood, with InfluxDB 2.0, all services will be on the same binary. It means that, the new sizing of InfluxDB will be, as maximum, the resources needed to run InfluxDB and the resources used by Kapacitor, with the danger think that a single query/task can kill InfluxDB, Kapacitor, Chronograf and Telegraf config provider, so:
- Are there some benchmark in order to know the new InfluxDB 2.0 performance?
- Are there new and enhanced controls to avoid the lost of entire TICK with a simple query/task?
- Are there new sizing guides in order to start thinking about asking for resources in our environment?