Skip to main content

Infrastructure

1. Monitoring and Observability​

Overview of Monitoring and Observability​

  • We’ve implemented a comprehensive monitoring and observability framework across our system to ensure high reliability, rapid incident response, and insights into system health.
  • Metrics, logs, and traces have been centralized to provide a holistic view of system performance and potential bottlenecks. This approach allows us to continuously monitor resource utilization, detect anomalies, and trace user requests for deeper insights.

2. Setting Up Monitoring Tools​

AWS CloudWatch​

  • CloudWatch Integration for EC2 and Other AWS Services: We have integrated AWS CloudWatch to monitor key performance metrics, including CPU utilization, network traffic, and disk usage for EC2 instances. Detailed monitoring is enabled for critical resources to capture fine-grained performance data. Custom metrics were also defined to monitor application-specific parameters.
  • Alarms: CloudWatch Alarms have been set up to trigger notifications when key thresholds are exceeded, such as CPU usage surpassing 80%. These alarms are linked to automated actions and notifications to minimize manual intervention.
  • AWS Cloudwatch setup guide

Uptime Robot​

  • Uptime Robot has been configured to monitor the uptime of our external-facing services. We’ve set up HTTP monitors to check the availability of our website and API endpoiints at five-minute intervals. Alerts are routed through integrations with Slack and email for real-time notifications in case of downtime.
  • Uptime robot setup guide

AWS SNS for Notifications​

  • For critical alarms and incidents, AWS SNS has been configured to send notifications via multiple channels, including email and SMS. We’ve created a central notification topic, allowing us to manage subscribers efficiently and ensuring key team members are informed of any issues immediately.

  • AWS SNS setup guide