Best Content on Cloud Technologies

AWS CloudWatch - Essentials

Managed Services is the one of the best features of cloud computing and in this article, we will take a look at the logging service by AWS which is called CloudWatch.

Logging for applications is essential for monitoring and security. CloudWatch can work as a central point for all of your logging requirements in the AWS cloud ecosystem. When you build complex apps that use multiple AWS services, you get lot of benefits by using CloudWatch by integrating it into your apps instead of relying on, other, isolated logging systems.

When it comes to logging, there are so many free and open source options and also SaaS offering that can create central logging. All of these offerings have their own pros and cons, but if you are using AWS, I would definitely recommend you take a look at CloudWatch before you evaluate other solution.

Some of the cool things that you can do with CloudWatch are.

  1. Monitor your applications by streaming the logs to CloudWatch.
  2. Create custom log groups to facilitate the timestamp-based logging or classification based on business need.
  3. Stream logs from your Lambda (Serverless) applications.
  4. Query the logs using a variety of options for debugging requirements.
  5. Store and archive logs to S3 or Glacier for compliance requirements.
  6. You can organise your logs in “namespaces” .
  7. You can use Metric to store a time-based data point and you can create custom metrics.
  8. Use Alarms to trigger actions i.e. other services. This is one of the powerful feature of CloudWatch.

CloudWatch concepts at high level

  1. Namespaces - Organise your logs into named spaces using Namespaces.
  2. Metrics - A smallest unit of data point with time information. You should identify the metrics to build reports if you are building custom metrics.
  3. Dimensions - Add more details to your metric to be able to filter your metrics. For example, every time you restart a service in your VM, use dimensions like ServiceName and ServiceRestartCount to understand the behaviuor.
  4. Resolution - Means, how frequently you log your data. From minute to second. Remember that, logging for every second costs more than, well, minute.
  5. Statistic, Units, Periods, Aggregation, Percentile - These are all used to build more fine-grained logs to build most robust monitoring system.
  6. Alarms - This is one of the most important feature.

Conclusion

You can use CloudWatch to build a powerful monitoring and alert system for your business-critical applications.