#!/bin/bash -e # run empty graph, with prometheus support $timeout --kill-after=60s 55s "$MGMT" run --tmp-prefix --prometheus --prometheus-listen :52637 & pid=$! sleep 5s # let it converge # Check that etcd metrics are loaded curl 127.0.0.1:52637/metrics | grep "^etcd_server_has_leader 1" # Check that go metrics are loaded curl 127.0.0.1:52637/metrics | grep "^go_goroutines [0-9]\+" killall -SIGINT mgmt # send ^C to exit mgmt wait $pid # get exit status exit $?