ams-managed-public-cloud

Deploy Azure Cosmos DB Apache Gremlin monitoring

This terraform script will deploy the following monitoring alerts for a Cosmos DB Apache Gremlin

  1. Resource Health
    Resource Health status is no longer ‘Available’.
    This indicates potential resource constraints impacting operational efficiency and service availability.
    The alert priority level is 2

  2. Region Failed
    Whenever the count Region Failed Over is greater than 0.
    This indicates instances where a region has experienced a failover event, potentially affecting service availability and requiring attention to ensure seamless operation.
    The alert priority level is 3

  3. Total Requests
    Whenever the count Total Requests is greater or less than dynamic criteria.
    This alerts to situations where there’s a significant variation in the number of requests, which may indicate changes in user activity or system behavior.
    The alert priority level is 5

  4. Normalized RU Consumption High
    Whenever the maximum Normalized RU Consumption* is greater than 90%.
    This indicates that database resources are nearing their maximum capacity, potentially impacting performance and scalability.
    The alert priority level is 3

  5. Normalized RU Consumption Critical
    Whenever the maximum Normalized RU Consumption* is greater than 98%.
    This signifies a critical situation where database resources are severely strained, likely leading to performance degradation and service disruptions.
    The alert priority level is 2

  6. Service Availability
    Whenever the average Service Availability is less than 100%.
    This alerts to instances where the service may experience downtime or interruptions, affecting user access and business continuity.
    The alert priority level is 2

  7. Region Offlined
    Whenever the count Region Offlined is greater than 0.
    This indicates situations where a region has become unavailable, potentially disrupting service availability and requiring immediate attention.
    The alert priority level is 3

  8. Gremlin Database Deleted
    Whenever the count Gremlin Database Deleted is greater than 0.
    This indicates instances where a Gremlin database has been intentionally removed, potentially impacting data availability and requiring investigation to ensure proper data management.
    The alert priority level is 2

    *Normalized RU Consumption refers to the amount of Request Units (measured in Request Units per second) consumed by a database operation, adjusted to a standardized scale for comparison across different databases and workloads.