linkerd-prometheus configuration #4892
-
Is it possible to configure linkerd-prometheus to pull metrics from apps beside linkerd-proxy. I want to set config settings like ports and endpoints with basic auth headers so I can use one prometheus service in my cluster that comes with linkerd. Thus I will be able to see both app and network metrics together. Is it possible? Thanks. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 1 reply
-
I'd recommend setting up a separate Prometheus instance for your use case. The data stored in the Linkerd Prometheus is retained for up to only 6 hours, and on restart/upgrade, they get wiped. As of the more recent edge releases, the Linkerd Prometheus is now an optional add-on. You can disable it following the instructions here. I think the exact flag to disable to Prometheus add-on is |
Beta Was this translation helpful? Give feedback.
-
Although this response may be late, it could still be valuable to someone. I recently encountered a situation where I required additional metrics for Canary deployment with Flagger. Interestingly, you can generate additional targets for To accomplish this, create an extra This configuration will add additional targets in prometheus.yaml file at this location.
|
Beta Was this translation helpful? Give feedback.
I'd recommend setting up a separate Prometheus instance for your use case. The data stored in the Linkerd Prometheus is retained for up to only 6 hours, and on restart/upgrade, they get wiped.
As of the more recent edge releases, the Linkerd Prometheus is now an optional add-on. You can disable it following the instructions here. I think the exact flag to disable to Prometheus add-on is
prometheus.enable=false
. It's in thevalues.yaml
file. Then set up your own Prometheus to scrape Linkerd metrics following this doc.