add model serving metrics (#18881)

* add model serving metrics * Update databricks/metadata.csv Co-authored-by: Steven Yuen <[email protected]> * Update manifest.json * add dashboard and monitors * fix format * fix format * fix format * add gpu monitors, add more metrics to csv, finished dashboard * json * json * fix monitors * fix metric to check * fix metric to check * add to dashboard * change served model name to served entity name * change 4xx and 5xx count to rate * rename 4xx and 5xx error count to rate * error rates are per min * Update metadata.csv * fix monitors * final dashboard version, add line to overview * fix databricks_cost_overview dashboard name * rearrange widgets --------- Co-authored-by: Steven Yuen <[email protected]>
DataDog · Nov 22, 2024 · 76ff273 · 76ff273
1 parent af4f594
commit 76ff273
Show file tree

Hide file tree

Showing 12 changed files with 1,434 additions and 2 deletions.
diff --git a/databricks/README.md b/databricks/README.md
@@ -19,6 +19,7 @@ Datadog offers several Databricks monitoring capabilities.
 
 [Infrastructure Monitoring][28] gives you a limited subset of the Data Jobs Monitoring functionality - visibility into the resource utilization of your Databricks clusters and Apache Spark performance metrics.
 
+Model serving metrics provide insights into how your  Databricks model serving infrastructure is performing. With these metrics, you can detect endpoints that have high error rate, high latency, are over/under provisioned, and more.
 ## Setup
 
 ### Installation

diff --git a/databricks/assets/dashboards/databricks_cost_overview.json b/databricks/assets/dashboards/databricks_cost_overview.json
@@ -879,4 +879,4 @@
     "layout_type": "ordered",
     "notify_list": [],
     "reflow_type": "fixed"
-  }
+}