ClaimedAccountConcurrency is the AWS Lambda metric that accurately reflects how much of your regional concurrency pool is unavailable for new invocations — unlike ConcurrentExecutions, which ignores reserved and provisioned capacity. This guide explains how Lambda concurrency works, why ClaimedAccountConcurrency is the correct metric to monitor, and walks through setting up a CloudWatch alarm at 70% utilization. It also covers deploying a CDK-based dashboard for regional visibility and automating Service Quotas increase requests via a Lambda function triggered directly by the alarm. Key caveats include the burst scaling rate limit, the difference between reserved and provisioned concurrency, and the fact that quota increases require manual AWS review and are not instant.
