Monitor across virtual machines within a cloud

Problem

The ability to monitor VM instances in a cloud is essential for optimizing instance performance, better allocation of resources, resolving issues around instance health, right-sizing, load-balancing and VM pools and managing service expectations.

Solution

Enable the monitoring function as a service on the cloud. The service should provide access to launch cloud monitors, authenticate itself with instances, authorize users, assign instances to monitors, provide ability to set the level of monitoring and the desired metrics to be gathered. The as-a-service design should be able to interface with cloud load balancing and auto-scale services. The service should support access through direct and proxy interfaces.

Click to Enlarge

Application

• Visibility into resource utilization, operational performance, demand patterns and optimize auto-scaling and VM pooling.
• Real time event detection and triaging threats and issues

Impacts

• Performance and resource impacts depending on extent of information gathered and number of VM instances monitored by the service
• Data traverses public network

Anti-Patterns

• Custom monitoring tools based on technology stack used with some level of manual handling of alerts and notifications
• Agent-based monitoring. Pull-based reporting from log repositories. Dashboard-based alerts and notifications for manual processing

Related Patterns

• As-a-Service Pattern
• Hardware device in software form Pattern

References

• AWS CloudWatch
• NimSoft