Give a solution to fix this. [9:15 AM] @data-engineering H...

Prompt

Give a solution to fix this. [9:15 AM] @data-engineering Hey team, we're getting reports from the Acme Corp tenant (prod) that their spend data pipeline hasn't updated since Friday 6 PM. Their dashboard is showing stale metrics and the CFO called our account manager asking wtf is going on. I checked Airflow and there are a bunch of failed DAG runs in the etl_spend_aggregation and client_metrics_transform jobs. The logs are pretty unclear—something about connection timeouts but I'm not sure if that's the API, the database, or the Lambda layer timing out. One of the devs mentioned they pushed something to the data-ingestion service on Friday around 4 PM but I don't have details on what changed. Could be nothing, could be everything. Also, I'm seeing weird Lambda cold starts in CloudWatch—some invocations taking 12+ seconds. Not sure if that's related to the pipeline failure or if it's just the usual AWS nonsense. The Acme Corp tenant is one of our biggest clients for the DSO Agent product, so this is kind of urgent. Can someone pull the logs, check what changed in the deployment, and figure out if we need to rollback or if it's something else? I'm gonna grab coffee. Let me know what you find.