There are two key ways you can respond to incidents and events when operating this Guidance. First, deviations in the data invoke a Lambda function, which then filters the severity and, if necessary, invokes an Amazon SNS email or SMS text notification to your operations team. Second, feedback is incorporated into the SageMaker machine learning model in the form of additional training data. This improves the quality of the model by constantly adjusting the threshold for SATCOM data anomalies.

Read the Operational Excellence whitepaper

A number of design decisions were factored for people and machine access when deploying this Guidance. For one, the principle of least privilege is applied with all services using the minimal IAM role permissions to process their function. Additionally, the AWS Shared Responsibility Model dictates that for AWS managed services, it is the responsibility of AWS to ensure the confidentiality, integrity, and the availability of its services. Meaning, those who use this Guidance do not have the responsibility to protect and maintain the underlying compute and network for the services used (such as Lambda and QuickSight). AWS implements the appropriate controls and maintenance as outlined by its internal polices and the many regulatory, legal, and compliance frameworks. Moreover, server-side encryption through AWS KMS is applied to the Kinesis Data Firehose records. Finally, the post-processed SATCOM analytics data-lake leverages Amazon S3, which automatically encrypts all new objects added to buckets on the server side, using AES-256 by default.

Read the Security whitepaper

This architecture leverages serverless tooling which auto-scales up (or down) on demand, helping you implement a highly available network topology. In addition, Lambda runs functions in multiple Availability Zones to ensure that it is available to process events in the case of a rare, but possible, service interruption in a single zone. The Lambda functions are also stateless, enabling as many copies of the function as needed to scale to the rate of incoming events.  Finally, the Amazon S3 Standard storage class is designed for 99.99% availability.

Read the Reliability whitepaper

The AWS services used throughout this Guidance were selected to provide a centralized and managed capability that ensures you can efficiently scale your implementation to any number of teleports and configurations.  In addition, the data partitioning in Kinesis Data Firehose and the Amazon S3 data lake enable highly efficient queries in Athena downstream. Also, consider reviewing the associated blog Creating satellite communications data analytics pipelines with AWS serverless technologies. It can help you get started through a code repository and AWS CloudFormation templates.

Read the Performance Efficiency whitepaper

This Guidance scales to continually match demand while ensuring that only the minimum resources are required in two primary ways. One, the control plane analytics pipeline leverages AWS serverless components; therefore, costs are only incurred when jobs are run. Second, the Guidance can scale up and down (to zero) as needed. As the number of satellite teleports and remote terminals grow, the architecture will adapt to match the demand, such as auto scaling the number of workers for an AWS Glue job.

Read the Cost Optimization whitepaper

This Guidance provides two benefits to help you meet your sustainability commitments. The first is that the IT infrastructure is elastic, which scales up and down based on usage, and does not provide excess compute resources, creating unintended emission.  You can follow your CO2 emissions using the AWS Sustainability Tools. The second gain is through the agility brought to engineering teams, where technologies like AWS Glue and QuickSight can help you to optimize your engineering operations by increasing efficiency and minimizing emissions.

Read the Sustainability whitepaper