Hi everyone, Earlier this month the subject of performance monitoring with grafana came up. I for one would be very happy to see this implemented in Radiator in order to have the logs stored in a consolidated manner (assuming the database supports clustering), and readable by graphing applications.
Heikki's proof of concept for InfluxDB (posted February 2016 in the mailing list), which opens a separate socket for each message, looks like a good start. I would like to suggest that a log handler for InfluxDB should be written, and if so that it allows for batching multiple data points instead of writing one line per socket. This should provide much better performance according to Influx's website. Link to suggestion: https://docs.influxdata.com/influxdb/v0.13/guides/writing_data/
_______________________________________________ radiator mailing list radiator@open.com.au http://www.open.com.au/mailman/listinfo/radiator
