I am using telegraf on a box containing a bunch of services using prometheus metrics. I think it could be a good idea to aggregate all those metrics through telegraf before exposing them as a single http endpoint using the output prometheus.
Basically, I will use the push API from my services to target the telegraf pushgateway endpoint. This will avoid doing complex stuff as files or named pipes.
Looking at this comment it seems that I am not the only one trying to achieve this.
Thoughts?
This is a good idea, but there are a few things in flight for prometheus that need to be coordinated. I need to finish #3670 so that we can merge #5767, at that point we can create the Prometheus parser as mentioned in #4414 (perhaps we would need two parsers, one for text and one for protobuf).
After all this it should be fairly easy to add a pushgateway input.
What's going on with this issue?
I think the next step is to extract the parser from the prometheus input into a standalone Parser.
Most helpful comment
What's going on with this issue?