Add some useful promql queries.

google_cloud_logging_queries/kubernetes/oom

@@ -0,0 +1 @@
+"OOMKilling"

google_cloud_logging_queries/vertex_ai_agent_builder/get_requests_for_engines

@@ -0,0 +1 @@
+textPayload=~"HTTP Request: POST https://us-discoveryengine.googleapis.com/[^\s/]+/projects/[^\s/]+/locations/us/collections/[^\s/]+/engines/[^\s/]+/"

metrics_explorer/over_provisioned/cloud_run_job/notes.org

@@ -0,0 +1,21 @@
+* Metrics Explorer PromQL Queries
+** Cloud Run Job Memory Utilization
+Line is a percent of memory usage.
+#+begin_src text
+  histogram_quantile(0.99,sum by (job_name,le)(increase(run_googleapis_com:container_memory_utilizations_bucket{monitored_resource="cloud_run_job"}[${__interval}])))
+#+end_src
+** Memory usage bytes
+#+begin_src text
+  histogram_quantile(0.50,sum_over_time(run_googleapis_com:container_memory_usage_bucket{monitored_resource="cloud_run_job"}[${__interval}]))
+#+end_src
+** Memory allocation
+#+begin_src text
+  sum by (job_name)(rate(run_googleapis_com:container_memory_allocation_time{monitored_resource="cloud_run_job"}[${__interval}]))
+#+end_src
+** Difference
+#+begin_src text
+  sum by (job_name)(
+      rate(run_googleapis_com:container_memory_allocation_time{monitored_resource="cloud_run_job"}[${__interval}])
+      - rate(run_googleapis_com:container_memory_usage_bucket{monitored_resource="cloud_run_job"}[${__interval}])
+      )
+#+end_src

metrics_explorer/pod_memory_limit/promql

@@ -0,0 +1 @@
+sum by (pod_name)(avg_over_time(kubernetes_io:container_memory_limit_bytes{monitored_resource="k8s_container",container_name=~".*airflow-worker.*"}[${__interval}]))