There are many types of activities you might perform to support the overall deployment production line that aren’t a part of the actual pipeline itself. Some examples include logging aggregation, application monitoring, system monitoring and governance systems.
For logging aggregation, you might incorporate CloudWatch logging and/or tools like Loggly or Splunk. For application monitoring, there are many tools from which to choose including New Relic and Boundary and AppDynamics. For systems monitoring, there are tools like CloudWatch, Nagios and Zenoss.
There are many ways to achieve manage governance on when implementing cloud delivery. Governance is a much broader topic that I will cover in this course, however, the key thing to consider is to use automation systems to enable better governance. In other words, humans still create all the rules, but leave it up to automation systems to ensure compliance to the rules. One example of this is how your teams might choose to terminate AWS resources. On most projects at Stelligent, we have a rule that indicating that all non-canonical resources should be terminated after they’ve been in use for 48 hours. This means that when someone provisions – say – an EC2 instance, we’ve enabled systems to automatically terminate that resource. In doing this, we’ve implemented an open-source tool called CloudPatrol that helps us manage resource usage on our projects. Let’s have a look at CloudPatrol.
As you’ve seen, there are many more activities you’ll need to consider in creating a cloud-delivery system to support the pipeline as part of your overall systems infrastructure.