Fix 503 http error when handling incoming webhook requests (#1747)

Incoming webhook request especially involving GitHub Apps with many
installations take minutes, so the existing http handler timeout of 10s
was causing 503 failures. Once increased, OpenShift deployments were
failing with http 504 (Gateway Time-out) because of the default 30s
timeout on the pipelines-as-code-controller route.

config/openshift/10-routes.yaml

@@ -15,6 +15,8 @@
 apiVersion: route.openshift.io/v1
 kind: Route
 metadata:
+  annotations:
+    haproxy.router.openshift.io/timeout: 600s
   labels:
     app.kubernetes.io/component: controller
     app.kubernetes.io/instance: default

pkg/adapter/adapter.go

@@ -30,6 +30,11 @@ import (
 
 const globalAdapterPort = "8080"
 
+// For incoming webhook requests and GitHub Apps with many installations the handler takes long
+// e.g GitHub App with ~400 installations, it takes ~180s. For OpenShift deployments this also
+// requires matching timeout on the pipelines-as-code-controller route (default is 30s).
+const httpTimeoutHandler = 600 * time.Second
+
 type envConfig struct {
 	adapter.EnvConfig
 }
@@ -87,7 +92,7 @@ func (l *listener) Start(ctx context.Context) error {
 	srv := &http.Server{
 		Addr: ":" + adapterPort,
 		Handler: http.TimeoutHandler(mux,
-			10*time.Second, "Listener Timeout!\n"),
+			httpTimeoutHandler, "Listener Timeout!\n"),
 	}
 
 	enabled, tlsCertFile, tlsKeyFile := l.isTLSEnabled()