Improve CCT error logging

Signed-off-by: Stefan Büringer [email protected]

controllers/remote/cluster_cache_tracker.go

@@ -183,7 +183,7 @@ func NewClusterCacheTracker(manager ctrl.Manager, options ClusterCacheTrackerOpt
 func (t *ClusterCacheTracker) GetClient(ctx context.Context, cluster client.ObjectKey) (client.Client, error) {
 	accessor, err := t.getClusterAccessor(ctx, cluster)
 	if err != nil {
-		return nil, err
+		return nil, errors.Wrapf(err, "failed to get client")
 	}
 
 	return accessor.client, nil
@@ -198,7 +198,7 @@ func (t *ClusterCacheTracker) GetReader(ctx context.Context, cluster client.Obje
 func (t *ClusterCacheTracker) GetRESTConfig(ctc context.Context, cluster client.ObjectKey) (*rest.Config, error) {
 	accessor, err := t.getClusterAccessor(ctc, cluster)
 	if err != nil {
-		return nil, err
+		return nil, errors.Wrapf(err, "failed to get REST config")
 	}
 
 	return accessor.config, nil
@@ -208,7 +208,7 @@ func (t *ClusterCacheTracker) GetRESTConfig(ctc context.Context, cluster client.
 func (t *ClusterCacheTracker) GetEtcdClientCertificateKey(ctx context.Context, cluster client.ObjectKey) (*rsa.PrivateKey, error) {
 	accessor, err := t.getClusterAccessor(ctx, cluster)
 	if err != nil {
-		return nil, err
+		return nil, errors.Wrapf(err, "failed to get etcd client certificate key")
 	}
 
 	return accessor.etcdClientCertificateKey, nil
@@ -267,7 +267,7 @@ func (t *ClusterCacheTracker) getClusterAccessor(ctx context.Context, cluster cl
 	// for the cluster at the same time.
 	// Return an error if another go routine already tries to create a clusterAccessor.
 	if ok := t.clusterLock.TryLock(cluster); !ok {
-		return nil, errors.Wrapf(ErrClusterLocked, "failed to create cluster accessor: failed to get lock for cluster")
+		return nil, errors.Wrapf(ErrClusterLocked, "failed to create cluster accessor: failed to get lock for cluster (probably because another worker is trying to create the client at the moment)")
 	}
 	defer t.clusterLock.Unlock(cluster)
 
@@ -462,12 +462,6 @@ func (t *ClusterCacheTracker) createCachedClient(ctx context.Context, config *re
 		cancelFunc: cacheCtxCancel,
 	}
 
-	for _, index := range t.indexes {
-		if err := cache.IndexField(ctx, index.Object, index.Field, index.ExtractValue); err != nil {
-			return nil, errors.Wrapf(err, "error creating cached client for remote cluster %q: error adding index for field %q to cache", cluster.String(), index.Field)
-		}
-	}
-
 	// Create the client for the remote cluster
 	cachedClient, err := client.New(config, client.Options{
 		Scheme:     t.scheme,
@@ -494,6 +488,12 @@ func (t *ClusterCacheTracker) createCachedClient(ctx context.Context, config *re
 		return nil, fmt.Errorf("failed waiting for cache for remote cluster %v to sync: %w", cluster, cacheCtx.Err())
 	}
 
+	for _, index := range t.indexes {
+		if err := cache.IndexField(ctx, index.Object, index.Field, index.ExtractValue); err != nil {
+			return nil, errors.Wrapf(err, "error creating cached client for remote cluster %q: error adding index for field %q to cache", cluster.String(), index.Field)
+		}
+	}
+
 	// Wrap the cached client with a client that sets timeouts on all Get and List calls
 	// If we don't set timeouts here Get and List calls can get stuck if they lazily create a new informer
 	// and the informer than doesn't sync because the workload cluster apiserver is not reachable.
@@ -574,7 +574,7 @@ func (t *ClusterCacheTracker) Watch(ctx context.Context, input WatchInput) error
 	// We have to lock the cluster, so that the watch is not created multiple times in parallel.
 	ok := t.clusterLock.TryLock(input.Cluster)
 	if !ok {
-		return errors.Wrapf(ErrClusterLocked, "failed to add %T watch on cluster %s: failed to get lock for cluster", input.Kind, klog.KRef(input.Cluster.Namespace, input.Cluster.Name))
+		return errors.Wrapf(ErrClusterLocked, "failed to add %T watch on cluster %s: failed to get lock for cluster (probably because another worker is trying to create the client at the moment)", input.Kind, klog.KRef(input.Cluster.Namespace, input.Cluster.Name))
 	}
 	defer t.clusterLock.Unlock(input.Cluster)