fix: Update uptime metrics query and fix node pool exception (#1200)

yuna-tzeng · web-flow · commit 53884d16bc9b · 2026-03-03T16:29:20.000+08:00
- In `query_uptime_metrics`, replace `query_time_series` with `list_time_series` to align with the recent `gcloud_util.py` refactoring.
- Fix a syntax error in the exception handling within `node_pool_util`.
diff --git a/dags/tpu_observability/utils/node_pool_util.py b/dags/tpu_observability/utils/node_pool_util.py
@@ -243,7 +243,8 @@ def create(
         "gcloud container operations list "
         f"--project={node_pool.project_id} "
         f"--region={node_pool.location} "
-        f"--format=json(name,status)"
+        f"--filter='status=RUNNING AND targetLink:{node_pool.node_pool_name}' "
+        f"--format='json(name,status)'"
     )
     debug_res = subprocess.run_exec(debug_cmd)
 

Original file line number	Diff line number	Diff line change
`@@ -243,7 +243,8 @@ def create(`
`243`	`243`	`"gcloud container operations list "`
`244`	`244`	`f"--project={node_pool.project_id} "`
`245`	`245`	`f"--region={node_pool.location} "`
`246`		`- f"--format=json(name,status)"`
	`246`	`+ f"--filter='status=RUNNING AND targetLink:{node_pool.node_pool_name}' "`
	`247`	`+ f"--format='json(name,status)'"`
`247`	`248`	`)`
`248`	`249`	`debug_res = subprocess.run_exec(debug_cmd)`
`249`	`250`