Posted On: Mar 27, 2024
Starting today, you can monitor your AWS Global Accelerator resources using five new Amazon CloudWatch metrics: ActiveFlowCount, PacketsProcessed, TCP_AGA_Reset_Count, TCP_Client_Reset_Count, and TCP_Endpoint_Reset_Count. With these metrics, you can monitor the total number of active TCP or UDP flows between your application’s clients and your accelerator endpoints, the total numbers of packets processed for an accelerator, and the total number of resets generated by Global Accelerator, the client, and the endpoint behind an accelerator. You can create CloudWatch alarms with the five new metrics to more quickly and easily detect issues with your accelerator endpoints.
These new metrics complement existing CloudWatch metrics for Global Accelerator, which include the following: the total number of incoming and outgoing bytes processed by an accelerator, the total number of new TCP or UDP flows between clients and application endpoints, and the total number of healthy endpoints and unhealthy endpoints for an accelerator. You can set alarms and specify automated actions with CloudWatch metrics, using defined thresholds. You can also build metric dashboards and view metrics directly in the CloudWatch console for your accelerators, listeners, and endpoint groups. To learn more about CloudWatch metrics for Global Accelerator, see the Global Accelerator documentation.