Best Practice 18.1 – Monitor and manage your IoT service quotas using available tools and metrics
Monitoring enables you to be aware of which service limits you might be hitting, allowing you to engineer your application to cope with the hard limits or to request the increase of a soft limit with enough lead time.
Recommendation 18.1.1 – Familiarize yourself with the service limits of the different IoT service
-
Pay attention to which limits are soft limits and which are hard limits as they require different approaches.
-
For example:
-
A hard limit, such a control plane request rate, would require changes in the application behavior to avoid the event repeating too often. Workarounds for hard limits might require different design decisions, such as using multiple accounts. It’s good to know the hard limits in advance so that you can make these design decisions as early as possible in the development process.
-
Soft limits should be monitored to anticipate the need for additional capacity and provide sufficient notice so that a request for a limit increase can be made well ahead of time.
-
-
For more:
-
For example:
-
For AWS IoT Core alert on
RulesMessageThrottles,Connect.ClientIDThrottle,Connect.Throttle,PublishIn.Throttle,Subscribe.Throttle,Unsubscribe.Throttle -
For AWS IoT Analytics alert on
ActionExecutionThrottled,PipelineConcurrentExecutionCount -
For AWS IoT Device Management monitor Active continuous jobs, Active snapshot jobs in Service Quotas
-
For AWS IoT SiteWise Monitor the quotas in Service Quotas
-
-
For more:
