When you start using some AWS resources, it will become useful to monitor them as well. Thus, Amazon has provided the Cloud Watch service to efficiently accomplish this task. It follows the pattern of being a low entry cost tool that can grow to a valuable enterprise solution.
What Does Cloud Watch Do?
In the world of odd names for Amazon services, this one ignores the trend. The name sums up the service very well. This one helps you watch your cloud services on Amazon. The tool works in a fashion similar to other monitoring options on the market. It provides a way to build dashboards, watches, and send notifications. The tool makes it easy to select from a list of attributes to monitor for your various AWS services. Better yet, the pay-as-you-go model makes it easy to experiment with measures and thresholds that best suit your needs.
Warnings, Errors, and Logs (Oh My)
The value of a tool such as Cloud Watch is to be proactive about system issues as well as debug failures. Of course, the monitoring options can be configured to log data for debugging. However, the alert rules can send out warnings as resources approach failure instead of waiting for an error notification after a crash.
The temptation is to keep the settings simple and just send out a notification once a system is unavailable. Avoid this. Consider the warning signs that occur before complete failure. This may be heavy disk usage, little available memory, or reduced response times. The rules around an alarm can provide notifications about these trends. Thus, a failure can potentially be avoided.
Live And Learn
Monitoring tools are only as good as those that configure them. Proper use requires experience and knowledge of the services beyond a cursory level. This technology can seem intimidating to the novice. Luckily, experience is a great teacher in using these tools. Thus, the best approach for someone new to monitoring is to browse the values available and watch how they change. This method can be achieved by using a dashboard. As time marches on the highs and lows of the systems monitored will be easy to read. Then the trends and spikes can be reviewed to learn what portends a crash or need for more resources.
Amazon makes it easy to get started with most of the services they provide. This service is no different. So skip the excuses and start learning how to monitor your systems. The crash you avoid may be your own.
