# MIDACOST05-BP01 Implement a buffering or throttling approach
<a name="midacost05-bp01"></a>

 Implement balanced resource utilization that handles varying workload demands while maintaining cost efficiency for manufacturing systems. This includes prioritizing critical processes while queuing less time-sensitive tasks and implementing appropriate scaling triggers aligned with production cycles. 

 **Desired outcome:** Balanced resource utilization that handles varying workload demands while maintaining cost efficiency. 

 **Common anti-patterns:** 
+  Implementing throttling on time-critical manufacturing processes 
+  Using the same buffering strategy for all types of industrial data 
+  Overlooking real-time requirements of production monitoring systems 
+  Setting queue limits without considering production batch sizes 
+  Implementing aggressive throttling that impacts quality data collection 
+  Not accounting for upstream and downstream dependencies in manufacturing processes 
+  Using standard IT buffering patterns without adapting to manufacturing needs 

 **Benefits of establishing this best practice:** 
+  Controlled resource consumption 
+  Avoided system overload 
+  Optimized costs during peak periods 
+  Improved system stability 

 **Level of risk exposed if this best practice is not established:** Medium 

## Implementation guidance
<a name="implementation-guidance-65"></a>

 Before you begin, you will need: 
+  Documented critical and non-critical manufacturing processes 
+  Peak resource utilization patterns for different production phases 
+  Response time requirements for various manufacturing systems 

 Key decisions needed: 
+  Resource allocation priorities for critical vs. non-critical processes 
+  Throttling thresholds for different types of manufacturing workloads 
+  Queue configurations for deferrable processes 
+  Scaling triggers aligned with production cycles and peaks 

Implement buffering and throttling mechanisms to manage cloud resource utilization during manufacturing peaks. Design a system that prioritizes critical processes (for example, real-time monitoring, quality control) for immediate resource access, while queuing less time-sensitive tasks (for example, batch analytics, report generation). Use auto-scaling for baseline capacity but implement throttling to help prevent non-critical tasks from consuming resources needed for production-critical operations. 

 Consider the following: 
+  Using Spot Instances for interruptible, non-critical workloads 
+  Implementing reserved capacity for predictable, critical processes 
+  Using serverless technologies for sporadic, scalable tasks 

 Regularly review and adjust your buffering and throttling strategies based on changing production patterns and business needs. 

### Implementation steps
<a name="implementation-steps-45"></a>
+  Identify and categorize manufacturing workloads: 
  +  Critical real-time processes (for example, process control, safety systems) 
  +  Time-sensitive operations (for example, quality inspections, inventory updates) 
  +  Deferrable tasks (for example, long-term analytics, reporting) 
+  Design resource allocation strategies: 
  +  Priority-based access for critical systems 
  +  Queueing mechanisms for non-critical operations 
  +  Load balancing across production lines or facilities 
+  Implement OT-aware monitoring: 
  +  Set up real-time monitoring for critical production KPIs 
  +  Configure alerts based on manufacturing thresholds 
  +  Integrate with SCADA or MES for comprehensive visibility 
+  Establish OT-IT integrated scaling mechanisms: 
  +  Automatic scaling triggered by production volumes 
  +  Resource reservation for planned production increases 
  +  Gradual scale-down aligned with shift changes or maintenance windows 
+  Conduct regular performance and cost reviews: 
  +  Analyze resource utilization against production output 
  +  Identify opportunities for optimization without impacting OT 
  +  Adjust strategies based on changing manufacturing requirements 
+  Implement feedback loops with shop floor: 
  +  Gather input from operators on system performance 
  +  Align IT resource adjustments with production schedules 
  +  Continuously refine based on real-world manufacturing impact 

## Key AWS services
<a name="key-aws-services-27"></a>
+  Amazon SQS 
+  Amazon Kinesis 
+  AWS Auto Scaling 
+  Amazon API Gateway 

## Resources
<a name="resources-66"></a>

 **Related documents:** 
+  [Amazon Simple Queue Service](https://docs.aws.amazon.com/AWSSimpleQueueService/latest/SQSDeveloperGuide/welcome.html) 
+  [Amazon Kinesis Data Streams Developer Guide](https://docs.aws.amazon.com/streams/latest/dev/introduction.html) 
+  [AWS Auto Scaling](https://docs.aws.amazon.com/autoscaling/plans/userguide/what-is-aws-auto-scaling.html) 
+  [Throttle requests to your REST APIs for better throughput in API Gateway](https://docs.aws.amazon.com/apigateway/latest/developerguide/api-gateway-request-throttling.html)