Fix GRPC Retry Configuration

.github/workflows/ci.yml

@@ -17,12 +17,14 @@ on:
       - "release/**"
 
 env:
-    GO_VERSION: 1.25
+  GO_VERSION: 1.25
 
 jobs:
   test:
     name: "Test"
     runs-on: ubuntu-latest
+    env:
+      GRPC_GO_RETRY: "on"
     steps:
       - uses: actions/checkout@v5
         with:

server/server.go

@@ -287,7 +287,7 @@ func DetectRemoteChainID(url string) (flowgo.ChainID, error) {
 		grpc.WithDefaultServiceConfig(utils.DefaultGRPCServiceConfig),
 	)
 	if err != nil {
-		return "", err
+		return "",fmt.Errorf("could not connect to remote access node: %w", err)
 	}
 	defer func() { _ = conn.Close() }()
 	client := flowaccess.NewAccessAPIClient(conn)

utils/grpc.go

@@ -18,27 +18,18 @@
 
 package utils
 
-// DefaultGRPCServiceConfig provides automatic retry configuration for transient gRPC errors.
-// This config is applied to all remote gRPC connections to handle network flakiness in CI
-// and other environments.
-//
-// Retries on:
-// - UNAVAILABLE: Service temporarily unavailable (e.g., node restarting)
-// - RESOURCE_EXHAUSTED: Rate limiting from remote node
-// - UNKNOWN: Connection failures, DNS issues, and other network errors
-//
-// Note: We only retry on clearly transient network/availability errors.
-// We do NOT retry on INTERNAL (programming errors), ABORTED (conflicts),
-// or DEADLINE_EXCEEDED (to avoid cascading failures on slow services).
+// DefaultGRPCServiceConfig configures native gRPC retries for transient failures.
+// The empty object wildcard [{}] matches all services and methods.
 const DefaultGRPCServiceConfig = `{
 	"methodConfig": [{
-		"name": [{"service": ""}],
+		"name": [{}],
 		"retryPolicy": {
-			"maxAttempts": 5,
-			"initialBackoff": "0.1s",
+			"maxAttempts": 8,
+			"initialBackoff": "1s",
 			"maxBackoff": "30s",
 			"backoffMultiplier": 2,
 			"retryableStatusCodes": ["UNAVAILABLE", "RESOURCE_EXHAUSTED", "UNKNOWN"]
 		}
 	}]
 }`
+