You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Prometheus follows pull model but it provides PushGateway. While looking through the output plugins, I'm not sure which one is suitable to make it push-able from Telegraf output to Prometheus (through PushGateway)?
Maybe just need some more docs or a new feature request for that?
Proposal:
Telegraf output plugin -> Prometheus PushGateway?
Current behavior:
Currently, it's very convenient to use prometheus client output plugin to expose /metrics endpoint(s) for Prometheus Server to pull.
But exposing endpoints may not always be valid and push model might be more feasible, in some cases (e.g. in NAT topology).
Desired behavior:
We can simply enable an output plugin to push to PushGateway
Use case:
In NAT topology, using SNAT to access external world, including the independent multi-cluster Monitoring Control Plane is common. So we need to aggregate metrics by exposing one single entry point which, in Prometheus world, might be PushGateway.
The text was updated successfully, but these errors were encountered:
I think this issue would be addressed by adding the Prometheus serializer and pairing it with the http output. I'm going to close this issue and we have an issue open for the serializer in #4414.
Feature Request
Prometheus follows pull model but it provides PushGateway. While looking through the output plugins, I'm not sure which one is suitable to make it push-able from Telegraf output to Prometheus (through PushGateway)?
Maybe just need some more docs or a new feature request for that?
Proposal:
Telegraf output plugin -> Prometheus PushGateway?
Current behavior:
Currently, it's very convenient to use
prometheus client
output plugin to expose/metrics
endpoint(s) for Prometheus Server to pull.But exposing endpoints may not always be valid and push model might be more feasible, in some cases (e.g. in NAT topology).
Desired behavior:
We can simply enable an output plugin to push to PushGateway
Use case:
In NAT topology, using SNAT to access external world, including the independent multi-cluster Monitoring Control Plane is common. So we need to aggregate metrics by exposing one single entry point which, in Prometheus world, might be PushGateway.
The text was updated successfully, but these errors were encountered: