edit-icon download-icon

Metrics

Last Updated: Aug 17, 2018

Function Compute metrics are reported in three dimensions:

  • Region dimension

  • Service dimension

  • Function dimension

All metric values are aggregated to 1-minute granularity and are stored as timeseries data. The Duration metric is aggregated as average and all other metrics are aggregated as sum.

Region Dimension

Region level metrics are useful to measure and monitor the overall usage of FC resources in a region. The following table lists metrics and descriptions.

Region-dimension metric Unit Description
TotalInvocations Count Total number of invocations of the region, including all synchronous InvokeFunction API calls and requests handled by FC asynchronously. Note that asynchronous invocation metrics are only reported after FC finished handling the requests, not at the time asynchronous InvokeFunction API returns 202 (i.e. requests with HTTP status 202 are excluded).
BillableInvocations Count Number of invocations of the region that can be metered and billed, including the InvokeFunction requests that return the HTTP status 200 or the asynchronous invocation requests successfully processed by FC. Note that function errors such as syntax errors, handled/unhandled errors and function timeout are included in this metric.
Throttles Count Number of InvokeFunction requests of the region that are throttled, including the requests that call the InvokeFunction API for access and return the HTTP status 429 and the asynchronous invocation requests that fail to be executed due to throttling.
ClientErrors Count Number of InvokeFunction requests of the region that returned with the HTTP status 4xx (excluding 429) and the asynchronous invocation requests that fail to be executed due to client side errors.
ServerErrors Count Number of InvokeFunction requests of the region that returned the HTTP status 5XX and the asynchronous invocation requests that fail to be run due to server side (FC) errors.
BillableInvocationsRate % Percentage of BillableInvocations of the region from the total number of InvokeFunction requests of the region.
ThrottlesRate % Percentage of Throttles of the region from the total number of InvokeFunction requests of the region.
ClientErrorsRate % Percentage of ClientErrors of the region from the total number of InvokeFunction requests of the region.
ServerErrorsRate % Percentage of ServerErrors of the region from the total number of requests of the region.

Service Dimension

Service level metrics are useful to measure and monitor the overall usage of FC resources in a service. The following table lists metrics and descriptions.

Service-dimension metric Unit Description
TotalInvocations Count Total number of invocations of the service, including all synchronous InvokeFunction API calls and requests handled by FC asynchronously. Note that asynchronous invocation metrics are only reported after FC finished handling the requests, not at the time asynchronous InvokeFunction API returns 202 (i.e. requests with HTTP status 202 are excluded).
BillableInvocations Count Number of invocations of the service that can be metered and billed, including the InvokeFunction requests that return the HTTP status 200 or the asynchronous invocation requests successfully processed by FC. Note that function errors such as syntax errors, handled/unhandled errors and function timeout are included in this metric.
Throttles Count Number of InvokeFunction requests of the service that are throttled, including the requests that call the InvokeFunction API for access and return the HTTP status 429 and the asynchronous invocation requests that fail to be executed due to throttling.
ClientErrors Count Number of InvokeFunction requests of the service that returned with the HTTP status 4xx (excluding 429) and the asynchronous invocation requests that fail to be run due to client side errors.
ServerErrors Count Number of InvokeFunction requests of the service that returned the HTTP status 5XX and the asynchronous invocation requests that fail to be run due to server side (FC) errors.
BillableInvocationsRate % Percentage of BillableInvocations of the service from the total number of InvokeFunction requests of the service.
ThrottlesRate % Percentage of Throttles of the service from the total number of InvokeFunction requests of the service.
ClientErrorsRate % Percentage of ClientErrors of the service from the total number of InvokeFunction requests of the service.
ServerErrorsRate % Percentage of ServerErrors of the service from the total number of requests of the service.

Function Dimension

Function level metrics are useful to measure and monitor the usage of resources of a function. The following table lists the function-dimension metrics.

Function-dimension metric Unit Description
AvgDuration Millisecond The average time in milliseconds elapsed from the start of function execution to the stop during the aggregation period (1-minute). Unlike the metering rules that round up the request duration by 100ms, this metric reports actual function execution time.
MaxMemoryUsage MB The maximum memory in MB used by function executions during the aggregation period (1-minute).
TotalInvocations Count Total number of invocations of the function, including all synchronous InvokeFunction API calls and requests handled by FC asynchronously. Note that asynchronous invocation metrics are only reported after FC finished handling the requests, not at the time asynchronous InvokeFunction API returns 202 (i.e. requests with HTTP status 202 are excluded).
BillableInvocations Count Number of invocations of the function that can be metered and billed, including the InvokeFunction requests that return the HTTP status 200 or the asynchronous invocation requests successfully processed by FC. Note that function errors such as syntax errors, handled/unhandled errors and function timeout are included in this metric
FunctionErrors Count Number of function invocations encountered errors such as Handled/Unhandled/OOM/timeout that are caused by function execution.
Throttles Count Number of InvokeFunction requests of the function that are throttled, including the requests that call the InvokeFunction API for access and return the HTTP status 429 and the asynchronous invocation requests that fail to be executed due to throttling.
ClientErrors Count Number of InvokeFunction requests of the function that returned with the HTTP status 4xx (excluding 429) and the asynchronous invocation requests that fail to be run due to client side errors.
ServerErrors Count Number of InvokeFunction requests of the function that returned the HTTP status 5XX and the asynchronous invocation requests that fail to be run due to server side (FC) errors
BillableInvocationsRate % Percentage of BillableInvocations of the function from the total number of InvokeFunction requests of the function.
ThrottlesRate % Percentage of Throttles of the region from the function number of InvokeFunction requests of the function.
ClientErrorsRate % Percentage of ClientErrors of the function from the total number of InvokeFunction requests of the function.
ServerErrorsRate % Percentage of ServerErrors of the function from the total number of requests of the function.
Thank you! We've received your feedback.