CDP:Monitoring Integration Pattern
Centralization of Monitoring Tools
Problem to Be Solved
Monitoring (of services and resources, and the like) is necessary in system operation. A monitoring service is provided by the Amazon Cloud. However, because the monitoring service in the Amazon Cloud is unable to monitor the internal workings of a virtual server (the operating system, middleware, applications, and so forth), you need to have an independent monitoring system. This is a problem in that it results in multiple monitoring systems, and complicates the operations (especially the monitoring operations).
Explanation of the Cloud Solution/Pattern
The virtual server and the like are monitored by the AWS Cloud monitoring service, but you need to use your own system to monitor the operating system, middleware, applications, and so forth. The Cloud monitoring service provides an API, enabling you to use your monitoring system to perform centralized control, including of the Cloud side, through this API, to obtain information from the Cloud monitoring system.
Install monitoring software on the EC2 instance, to obtain monitoring information from the "CloudWatch" monitoring service of AWS.
- Install monitoring software (Nagios, Zabbix, Munin, or the like).
- Use a plug-in to obtain monitoring information using the CloudWatch API and to write that information to the monitoring software.
- Use the plug-in to perform monitoring, including the information from AWS.
- This lets you monitor the Cloud resources in the same way as monitoring of the operating system, middleware, and applications through your own system.
- While data is stored for two weeks in CloudWatch, you can extend the time period for storing data through acquiring the data using monitoring software.
- The use of monitoring software in relation to AWS resource status notifications let you customize the emailed reports, for example.
- An API usage fee is required for the monitoring software (plug-in) to obtain data periodically from CloudWatch through the API.