AWS Lambda Cheatsheet Collect these OOTB metrics with Datadog START YOUR FREE TRIAL

Key Performance Data

Note:
- Standard Cloudwatch metrics are available via Datadog's AWS integration
- Enhanced metrics are available for Lambda functions instrumented with the Datadog Lambda Library and Datadog Lambda Extension
- For more information on Lambda monitoring, visit https://dtdg.co/lambda-metrics

Standard Cloudwatch Metrics Enhanced Metrics from Datadog

The following are key Cloudwatch metrics for monitoring the health and performance of AWS Lambda functions. The following are Datadog's enhanced metrics for Lambda functions, which are available at second-level granularity and
provide deeper insight into performance and costs.
Function Utilization and Performance Metrics—more info
METRIC DESCRIPTION DATADOG METRIC NAME
METRIC DESCRIPTION AVAILABILITY METRIC NAME
Number of invocations in response to an event or an API call aws.lambda.enhanced.invocations
Execution time (ms) cloudwatch Duration
Elapsed time across init, runtime, and post-runtime durations aws.lambda.enhanced.duration
Memory allocated to a function (MB) logs Memory Size
Initialization time during a cold start (seconds) aws.lambda.enhanced.init_duration
Maximum memory used (MB) logs Max Memory Used
Execution time (seconds) aws.lambda.enhanced.runtime_duration
Maximum amount of memory used (MB) aws.lambda.enhanced.max_memory_used

Lambda Invocation Metrics—more info Estimated cost of an invocation ($) aws.lambda.enhanced.estimated_cost

Number of failed invocations caused by function errors aws.lambda.enhanced.errors
METRIC DESCRIPTION AVAILABILITY METRIC NAME
Number of function timeouts aws.lambda.enhanced.timeouts
Age of the last processed record in a stream's batch cloudwatch IteratorAge
Number of times a function runs out of memory aws.lambda.enhanced.out_of_memory
Number of invocations cloudwatch Invocations
Number of failed invocations cloudwatch Errors
Number of failed attempts to send an event to the cloudwatch DeadLetterErrors
dead-letter queue AWS Lambda Metadata
Datadog automatically enriches Lambda metric, log, and trace data with tags that provide
additional context for troubleshooting performance issues.

Lambda Concurrency Metrics—more info TAG NAME* DESCRIPTION DATA SOURCE

METRIC DESCRIPTION AVAILABILITY METRIC NAME aws_account AWS account name Enhanced metrics and logs

Number of functions processing events cloudwatch ConcurrentExecutions region Region where a function is deployed Enhanced metrics and logs
Number of functions processing events without cloudwatch UnreservedConcurrentExecutions functionname Function Name Enhanced metrics and logs
reserved concurrency
cold_start If the function experienced a cold start Enhanced metrics and logs
Number of invocations throttled due to concurrent cloudwatch Throttles
execution limits memorysize Memory allocated to a function (MB) Enhanced metrics and logs

executedversion Function version Enhanced metrics and logs

Lambda Provisioned Concurrency Metrics—more info runtime Function runtime (Python, Node.js, Ruby, Java, Go, .NET) Enhanced metrics and logs

architecture Function processor (arm64, x86_64) Enhanced metrics and logs

METRIC DESCRIPTION AVAILABILITY METRIC NAME
Number of functions processing events on cloudwatch ProvisionedConcurrentExecutions function_arn Function Amazon Resource Name (ARN) Traces
provisioned concurrency
status_code HTTP request status Traces
Percent of provisioned concurrency used cloudwatch ProvisionedConcurrencyUtilization
function_trigger Resource that triggered a function Traces
Number of invocations on provisioned concurrency cloudwatch ProvisionedConcurrencyInvocations
Number of concurrent invocations over the cloudwatch ProvisionedConcurrencySpilloverInvocations http.method** HTTP request method Traces
provisioned concurrency limit API Gateway URL that invoked the function Traces
http.url**
Percent of provisioned concurrency being used cloudwatch ProvisionedConcurrencyInvocations
* Tags are in addition to custom tags and Datadog's unified and integration tags
** Functions invoked by API Gateway
AWS Lambda Cheatsheet
Note:
- For metric descriptions and more information: https://dtdg.co/lambda-metrics

4. Errors
METRIC DESCRIPTION DATADOG METRIC NAME
Number of failed invocations caused by function aws.lambda.enhanced.errors
errors, grouped by function name
Number of invocations throttled due to concurrent aws.lambda.throttles
execution limits, grouped by function name

5. Memory
METRIC DESCRIPTION DATADOG METRIC NAME
Maximum amount of memory used (MB), grouped by aws.lambda.enhanced.max_memory_used
function name

1. Status overview
METRIC DESCRIPTION DATADOG METRIC NAME
Total number of invocations in response to an event or an API call aws.lambda.enhanced.invocations
Total number of failed invocations caused by function errors aws.lambda.enhanced.errors
Total number of invocations that were cold starts aws.lambda.enhanced.invocations
Estimated cost of an invocation ($) aws.lambda.enhanced.estimated_cost

2. Invocations, Cold Starts, and Provisioned Concurrency

METRIC DESCRIPTION DATADOG METRIC NAME
Total number of invocations in response to an event or an API call, aws.lambda.enhanced.invocations
grouped by function name
Total ​​number of invocations, grouped by function name aws.lambda.invocations
Estimated cost of an invocation ($), grouped by function name aws.lambda.enhanced.estimated_cost
Total number of invocations that were cold starts aws.lambda.enhanced.invocations
Total number of functions processing events on provisioned aws.lambda.provisioned_concurrent_executions
concurrency, grouped by function name
Percent of provisioned concurrency used, grouped by function name aws.lambda.provisioned_concurrency_utilization

3. Execution Time
METRIC DESCRIPTION DATADOG METRIC NAME
Elapsed time across init, runtime, and post-runtime durations aws.lambda.enhanced.duration
Execution time (p95), grouped by function name aws.lambda.duration.p95
Total number of time outs, grouped by function name aws.lambda.timeout

Monitor AWS Lambda Technology with Datadog Free START YOUR FREE TRIAL