LavinMQ::HTTP::MetricsServer
Serves the Prometheus metrics endpoints. One instance lives for the whole process: what it reports follows the node's role, so it never has to be closed and rebound when a follower is promoted to leader. It reports as a follower until a server is set with #leader=.
Constants
Log = LavinMQ::Log.for("metrics.server")
Constructors
new(amqp_server : LavinMQ::Server | Nil = nil, clustering_client : LavinMQ::Clustering::Client | Nil = nil)
SourceInstance methods
bind_tcp(address, port)
Sourceclose
SourceReport as a follower of the leader that the client replicates from
Report as leader, with the full set of metrics from the server
listen
Source