fix(snapshot-agent): return error from BuildDirectMemoryConfig on unparsable PID strings

When strconv.ParseInt fails for an entry in pidStrings, return an error rather than silently skipping the entry. Update BuildDirectMemoryConfig signature to return (*pb.BackendConfig, error) and propagate any config-building error in server Snapshot and Restore handlers.

TAG=agy

CONV=9fda09b9-fe6d-48bc-b524-2d232e0cdd4b

Signed-off-by: edwinhr716 <edandres249@gmail.com>

Author: Edwinhr716

pkg/snapshot-agent/backends/direct_memory.go

@@ -148,18 +148,20 @@ func ExtractDirectMemoryPIDStrings(config *pb.BackendConfig) []string {
 }
 
 // BuildDirectMemoryConfig wraps PID strings into a DirectMemory BackendConfig.
-func BuildDirectMemoryConfig(pidStrings []string) *pb.BackendConfig {
+func BuildDirectMemoryConfig(pidStrings []string) (*pb.BackendConfig, error) {
 	pids := make([]int32, 0, len(pidStrings))
 	for _, s := range pidStrings {
-		if pid, err := strconv.ParseInt(s, 10, 32); err == nil {
-			pids = append(pids, int32(pid))
+		pid, err := strconv.ParseInt(s, 10, 32)
+		if err != nil {
+			return nil, fmt.Errorf("invalid PID string %q: %w", s, err)
 		}
+		pids = append(pids, int32(pid))
 	}
 	return &pb.BackendConfig{
 		Backend: &pb.BackendConfig_DirectMemory{
 			DirectMemory: &pb.DirectMemoryBackendConfig{
 				ExplicitTarget: &pb.ProcessTarget{Pids: pids},
 			},
 		},
-	}
+	}, nil
 }

pkg/snapshot-agent/backends/direct_memory_test.go

@@ -195,7 +195,10 @@ func TestDirectMemoryHealthCheck(t *testing.T) {
 
 func TestDirectMemoryConfigHelpers(t *testing.T) {
 	pids := []string{"100", "200"}
-	cfg := backends.BuildDirectMemoryConfig(pids)
+	cfg, err := backends.BuildDirectMemoryConfig(pids)
+	if err != nil {
+		t.Fatalf("BuildDirectMemoryConfig() unexpected error: %v", err)
+	}
 	extracted := backends.ExtractDirectMemoryPIDStrings(cfg)
 	if !reflect.DeepEqual(extracted, pids) {
 		t.Errorf("ExtractDirectMemoryPIDStrings() = %v, want %v", extracted, pids)
@@ -204,4 +207,9 @@ func TestDirectMemoryConfigHelpers(t *testing.T) {
 	if len(backends.ExtractDirectMemoryPIDStrings(nil)) != 0 {
 		t.Errorf("Expected nil when extracting from nil config")
 	}
+
+	_, err = backends.BuildDirectMemoryConfig([]string{"100", "not-a-pid"})
+	if err == nil {
+		t.Errorf("BuildDirectMemoryConfig() expected error for invalid PID string, got nil")
+	}
 }

pkg/snapshot-agent/server/server.go

@@ -162,11 +162,15 @@ func (s *Server) buildSnapshotFn(
 					return pidErr
 				}
 				var reqConfig *pb.BackendConfig
+				var cfgErr error
 				if backendType == backends.BackendDirectMemory {
-					reqConfig = backends.BuildDirectMemoryConfig(allPIDStrings)
+					reqConfig, cfgErr = backends.BuildDirectMemoryConfig(allPIDStrings)
 				} else {
 					reqConfig = backends.BuildCudaConfig(allPIDStrings)
 				}
+				if cfgErr != nil {
+					return fmt.Errorf("failed to build backend config for job %s: %w", jobID, cfgErr)
+				}
 				req := backends.Request{JobID: jobID, Config: reqConfig}
 				if err := backend.Snapshot(bgCtx, req); err != nil {
 					return fmt.Errorf("failed to snapshot job %s: %w", jobID, err)
@@ -241,11 +245,15 @@ func (s *Server) buildRestoreFn(
 				}
 				slog.InfoContext(bgCtx, "Restoring PIDs", "pids", pidStrings, "backend", backendType)
 				var reqConfig *pb.BackendConfig
+				var cfgErr error
 				if backendType == backends.BackendDirectMemory {
-					reqConfig = backends.BuildDirectMemoryConfig(pidStrings)
+					reqConfig, cfgErr = backends.BuildDirectMemoryConfig(pidStrings)
 				} else {
 					reqConfig = backends.BuildCudaConfig(pidStrings)
 				}
+				if cfgErr != nil {
+					return fmt.Errorf("failed to build backend config for job %s: %w", jobID, cfgErr)
+				}
 				return backend.Restore(bgCtx, backends.Request{JobID: jobID, Config: reqConfig})
 			}, nil
 		case backends.BackendAppEndpoint, backends.BackendAppChannel: