[ws-rollout] Add prometheus init check

When given a non-usable `--prometheus-url`, We start the
rollout without verifying if the prometheus is reachable or not. This
is a problem as we will be unable to get the metrics from prometheus
and hence the rollout will be reverted later causing unnecessary
time waste.

This can be prevented by performing a simple check to see if the
prometheus is reachable or not. `up` query is used instead of
key metrics as we can't be sure of their existence.

Signed-off-by: Tarun Pothulapati <[email protected]>

Author: Pothulapati

components/workspace-rollout-job/cmd/root.go

@@ -96,6 +96,13 @@ var rootCmd = &cobra.Command{
 			return err
 		}
 
+		// Check if prometheus is reachable
+		err = analysis.CheckPrometheusReachable(ctx, conf.prometheusURL)
+		if err != nil {
+			log.WithError(err).Fatal("init: prometheus is not reachable")
+			return err
+		}
+
 		prometheusAnalyzer, err := analysis.NewWorkspaceKeyMetricsAnalyzer(ctx, config, conf.prometheusURL, conf.targetPositivePercentage, 30305)
 		if err != nil {
 			log.WithError(err).Fatal("failed to create a prometheus client")

components/workspace-rollout-job/pkg/analysis/analyzer.go

@@ -6,6 +6,10 @@ package analysis
 
 import (
 	"context"
+	"time"
+
+	"github.com/prometheus/client_golang/api"
+	v1 "github.com/prometheus/client_golang/api/prometheus/v1"
 )
 
 type Decision int
@@ -21,3 +25,24 @@ type Analyzer interface {
 	// repeatedly to determine whether to move forward on the rollout or not.
 	MoveForward(ctx context.Context, clusterName string) (Decision, error)
 }
+
+func CheckPrometheusReachable(ctx context.Context, prometheusURL string) error {
+	if prometheusURL == "" {
+		return nil
+	}
+
+	client, err := api.NewClient(api.Config{
+		Address: prometheusURL,
+	})
+	if err != nil {
+		return err
+	}
+
+	v1Client := v1.NewAPI(client)
+	_, _, err = v1Client.Query(ctx, "up", time.Now())
+	if err != nil {
+		return err
+	}
+
+	return nil
+}