change

pkg/common/helper.go

@@ -192,13 +192,6 @@ func EngineErrorIsNonFatal(err error) bool {
 	return errorMatches(err, nonFatalEngineErrorsPartialMatch)
 }
 
-func EngineErrorNeedsIndefiniteRetry(err error) bool {
-	var fatalEngineErrorsPartialMatch = []string{
-		"lookup v3io-webapi: i/o timeout",
-	}
-	return errorMatches(err, fatalEngineErrorsPartialMatch)
-}
-
 func errorMatches(err error, substrings []string) bool {
 	if err != nil && len(err.Error()) > 0 {
 		for _, substring := range substrings {

pkg/dataplane/streamconsumergroup/claim.go

@@ -130,20 +130,13 @@ func (c *claim) fetchRecordBatches(stopChannel chan struct{}, fetchInterval time
 		func(attempt int) (bool, error, int) {
 			c.currentShardLocation, err = c.getCurrentShardLocation(c.shardID)
 			if err != nil {
-				// if the error is fatal and requires external resolution,
+				// if the error is not fatal (as network issue),
 				// we don't want to fail; instead, we will inform the user via a log
-				if common.EngineErrorNeedsIndefiniteRetry(err) {
-					c.logger.ErrorWith("Failed to get shard location. Will retry until successful",
-						"error", err,
-						"shard", c.shardID)
+				if common.EngineErrorIsNonFatal(err) {
 					// for this type of error, we always increment the attempt counter
 					// this ensures the smooth operation of other components in Nuclio
 					// we avoid panicking and simply wait for the issue to be resolved
-					return true, errors.Wrap(err, "Failed to get shard location"), 1
-				}
-
-				if common.EngineErrorIsNonFatal(err) {
-					return true, errors.Wrap(err, "Failed to get shard location due to a network error"), 0
+					return true, errors.Wrap(err, "Failed to get shard location due to a network error"), 1
 				}
 
 				// requested for an immediate stop