fix(asap-tools): pass ms-resource to the Alibaba cluster_data_exporter

The exporter's --data-type accepts node or ms-resource; the config value
msresource was passed through, so the container exited at startup.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Author: zzylol

asap-tools/experiments/experiment_utils/services/cluster_data_exporter.py

@@ -13,6 +13,10 @@
 from experiment_utils.providers.base import InfrastructureProvider
 
 
+# Experiment-config data_type -> the exporter's --data-type value.
+ALIBABA_DATA_TYPE_CLI_VALUES = {"node": "node", "msresource": "ms-resource"}
+
+
 class ClusterDataExporterService(BaseService):
     """
     Service for managing cluster_data_exporter via Docker.
@@ -259,7 +263,8 @@ def _build_docker_command(
 
         elif provider == "alibaba":
             if "data_type" in config:
-                cmd_parts.append(f"--data-type={config['data_type']}")
+                data_type = ALIBABA_DATA_TYPE_CLI_VALUES[config["data_type"]]
+                cmd_parts.append(f"--data-type={data_type}")
             if "data_year" in config:
                 cmd_parts.append(f"--data-year={config['data_year']}")
 

asap-tools/experiments/tests/test_cluster_data_exporter_service.py

@@ -45,6 +45,20 @@ def test_unknown_year_raises(self):
             service._validate_alibaba_data("msresource", 2020)
 
 
+class DockerCommandTest(unittest.TestCase):
+    def test_msresource_uses_exporter_cli_value(self):
+        # The exporter's clap enum spells it ms-resource; passing the config
+        # value through made the container exit before serving metrics.
+        service = ClusterDataExporterService(RecordingProvider(), 0, "/traces")
+        service.container_name = "cde"
+        cmd = service._build_docker_command(
+            {"provider": "alibaba", "data_type": "msresource", "data_year": 2022},
+            port=40000,
+            output_dir="/out",
+        )
+        self.assertIn("--data-type=ms-resource", cmd)
+
+
 class NodeCountTest(unittest.TestCase):
     def test_more_than_one_worker_node_is_rejected(self):
         service = ClusterDataExporterService(RecordingProvider(), 0, "/traces")