I think Telegraf is nice to get started, but is then missing some advanced features (e.g., redis caching for lookups, etc) It was originally developed to ingest data into InfluxDB and that is still the main case. But we want to do much more there. Another example: we are using the same OPC ua library as Telegraf, but we added much more code for exception handling and providing a smooth user flow (e.g., connecting to all the endpoints while patching DNS with IP until we get one that works). Last but not least, we used it a couple of years back when we were still working with InfluxDB and encountered a lot of weird bugs. That still leaves a bad impression to us