Design principles
The following are design principles for operational excellence in the cloud:
-
Based on the workload and your business objectives, identify connected mobility key performance indicators (KPIs). The connected mobility KPIs should be linked to metrics from all layers: business, application, data, security, and infrastructure. Once identified, use the connected mobility KPIs to work backwards and identify operational KPIs.
-
Implement end-to-end observability and a single pane view of the connected mobility workload status from vehicle to the cloud and full stack of the application: Proactively monitor end user experience in-vehicle.
-
Develop a testing framework to simulate production-like conditions: Site Reliability Engineering (SRE) team are responsible for reliability of the application amidst frequent updates from development teams. The SRE team works closely with the development team to create new features and stabilize production systems. One of the challenges faced by the SRE team is to simulate the vehicle while trying to test or debug a connected mobility feature or a production issue. Develop a virtual vehicle in the cloud that has capability to simulate various real-world conditions including load and Mobile Network Operator (MNO) failures.
-
Implement mechanisms to improve developer productivity enabling reduction in Time to Market (TTM), which is the time it takes to take connected mobility features from concept to launching it in the industry. Develop deployment templates and provide a developer portal to quickly spin up new development and lab environments.
-
Implement a multi-stakeholder runbook to effectively respond to production events. Connected mobility is a complex system with multiple stakeholders from end users to Mobile Network Operators, cloud hyperscalers, downstream data consumers, and OEMs. To reduce the Mean Time to Restore (MTTR), a runbook should be developed for all critical events. Eventually, automate runbooks to streamline operational processes, reduce human error, and improve efficiency, leading to faster incident resolution and enhanced operational excellence.
Definitions
Observability: Observability allows users to understand a system's state from its external output and take (corrective) action. Observable systems yield meaningful, actionable data to their operators, allowing them to achieve favorable outcomes (faster incident response, increased developer productivity) and less toil and downtime.
Best practices implementing observability include:
-
Monitor what matters: Start with the connected mobility business key performance indicators (KPIs) and work backwards to the application and infrastructure metrics.
-
Collect telemetry from all layers: Connected mobility has dependencies and interactions with vehicles, Mobile Network Operators, cloud providers, internet service providers, AWS Partners, and other components—both within and outside your control—that can impact your business outcomes. It is important that you have a holistic view of your entire workload.
-
Propagate context: Collect logs, metrics, and traces, and propagate transaction IDs, which helps to perform correlation, analysis, anomaly detection, dashboarding, and alarms.
Leading versus lagging indicators: Leading indicators are metrics that are used to measure future performance. For example, customer satisfaction and connected mobility feature usage metrics can be used to predict renewal rates, as a happy and engaged customer is more likely to renew the paid subscription. Similarly, quality metrics of the feature releases can be a leading indicator to predict the failure rate of the system.
Lagging indicators are metrics used to measure past performance for example connected mobility subscription renewal rates, Mean Time Between Failures (MTBF), and remote command latency. Lagging indicators provide valuable feedback on the effectiveness of past decisions and help identify areas for improvement. Both leading and lagging indicators are important for managing and measuring operational efficiency of connected mobility workloads.
