From b0ccf6f28d03a236d5c0088f31d2f0be59fbb58e Mon Sep 17 00:00:00 2001
From: Rohith Pariki <rohithpariki@gmail.com>
Date: Sun, 11 Oct 2026 03:36:27 +0530
Subject: [PATCH] fix: return error instead of panicking on execution engine
 start failure

This commit improves the reliability of the execution engine startup by returning an error up the stack to main instead of immediately panicking, allowing graceful handling of startup failures.
---
 cmd/airavata-server/main.go      | 5 ++++-
 internal/app/app.go              | 9 ++++++---
 internal/orchestration/engine.go | 8 ++++----
 internal/server/server_test.go   | 5 ++++-
 4 files changed, 18 insertions(+), 9 deletions(-)

diff --git a/cmd/airavata-server/main.go b/cmd/airavata-server/main.go
index 03120a7d54..326f5eb6c8 100644
--- a/cmd/airavata-server/main.go
+++ b/cmd/airavata-server/main.go
@@ -166,7 +166,10 @@ func runServer() error {
 		return fmt.Errorf("no workflow backend configured")
 	}
 
-	svcs := app.NewServices(cfg, gdb, repos, workflowBackend)
+	svcs, err := app.NewServices(cfg, gdb, repos, workflowBackend)
+	if err != nil {
+		return fmt.Errorf("build services: %w", err)
+	}
 
 	srv := &http.Server{
 		Addr:              cfg.HTTPAddr,
diff --git a/internal/app/app.go b/internal/app/app.go
index dd003bd503..d10f41c35e 100644
--- a/internal/app/app.go
+++ b/internal/app/app.go
@@ -159,7 +159,7 @@ func NewRepositories(db *gorm.DB) *Repositories {
 //
 // The repositories stay local: they are an implementation detail of the services above
 // them, and nothing outside this function has ever wanted one directly.
-func NewServices(cfg config.Config, db *gorm.DB, repos *Repositories, workflowBackend backend.Backend) *Services {
+func NewServices(cfg config.Config, db *gorm.DB, repos *Repositories, workflowBackend backend.Backend) (*Services, error) {
 	// Repositories.
 	users := repos.Users
 	groups := repos.Groups
@@ -204,7 +204,10 @@ func NewServices(cfg config.Config, db *gorm.DB, repos *Repositories, workflowBa
 
 	executionEngine := orchestration.NewExecutionEngine(stagingTasks, submissionTasks,
 		monitoringTasks, storages, clusterConfigs, processes, deployments, templates, batchStatus, statuses, workflowBackend)
-	executionEngine.StartEngine()
+	
+	if err := executionEngine.StartEngine(); err != nil {
+		return nil, err
+	}
 
 	return &Services{
 		Config: cfg,
@@ -248,5 +251,5 @@ func NewServices(cfg config.Config, db *gorm.DB, repos *Repositories, workflowBa
 			monitoringTasks, commandTasks, executionEngine),
 
 		ExecutionEngine: executionEngine,
-	}
+	}, nil
 }
diff --git a/internal/orchestration/engine.go b/internal/orchestration/engine.go
index edd4b2a32d..6c3cffb106 100644
--- a/internal/orchestration/engine.go
+++ b/internal/orchestration/engine.go
@@ -115,10 +115,10 @@ func NewExecutionEngine(dataStagingTasks *processrepo.DataStagingTaskRepository,
 	}
 }
 
-func (w *ExecutionEngine) StartEngine() {
+func (w *ExecutionEngine) StartEngine() error {
 	if w.orchestrator == nil {
 		slog.Info("Execution engine built without a workflow backend; not starting the workflow worker")
-		return
+		return nil
 	}
 
 	slog.Info("Starting execution engine...........")
@@ -131,9 +131,9 @@ func (w *ExecutionEngine) StartEngine() {
 	w.orchestrator.RegisterActivity(w.completeProcess)
 
 	if err := w.orchestrator.Start(ctx); err != nil {
-		slog.Error("Failed to start execution engine", "error", err)
-		panic("Could not start execution engine")
+		return err
 	}
+	return nil
 }
 
 // This is invoked when an email related to a batch job is received.
diff --git a/internal/server/server_test.go b/internal/server/server_test.go
index 021cad94e8..35afb0e652 100644
--- a/internal/server/server_test.go
+++ b/internal/server/server_test.go
@@ -112,7 +112,10 @@ func newHarness(t *testing.T) *harness {
 
 	repos := app.NewRepositories(gdb)
 	// passing nil for workflow backend to prevent test panic due to sqlite driver collision
-	svcs := app.NewServices(cfg, gdb, repos, nil)
+	svcs, err := app.NewServices(cfg, gdb, repos, nil)
+	if err != nil {
+		t.Fatalf("Failed to initialize services: %v", err)
+	}
 	return &harness{
 		t:    t,
 		db:   gdb,