Recent Cloudwatch Application Signals Announcement

I’ve been in IT and DevOps for well over a decade, using a variety of monitoring tools like Nagios, or SaaS products like NewRelic or Datadog. All of these solutions have their own pros and cons. CloudWatch has slowly been adding features to compete with these, but they always seemed to be piecemeal, with logs here, dashboard there. Nothing that felt holistic.  

Recently at re:Invent, AWS  announced improvements to Amazon CloudWatch Application Signals.I recently attended a webinar with a demo and walked through the new features. I was impressed. 

The new Application Signals features for monitoring SLO levels, drill downs through applications, and rapid troubleshooting. Finally a turn key view of applications without needing to stitch together different parts of cloudwatch manually with events or building custom dashboards. 

A highlight for me was being able to see a breach in SLO level. Then follow the service to X-Ray to see how it is affecting the whole application. Jump from there to logs and error messages. 

This has the potential to disrupt the industry for cloud monitoring, especially since you can monitor custom applications, not just EKS. This means you could use AWS to monitor all of your cloud environments regardless of the provider. 

Don’t get too excited yet. Unfortunately, the new features are in open preview in the following regions: US East (N. Virginia), US East (Ohio), Europe (Ireland), Asia Pacific (Sydney), and Asia Pacific (Tokyo). The announcement mentions US West (Oregon), but I don’t have access to the SLO or Service options there yet. 

In addition, if you want to monitor something besides EKS, it would need to be a custom setup. Though the hosts of the webinar mentioned that using Lambda Insights as an option would be coming soon. 

I’m looking forward to this service being released more widely to give it a test run on our applications. Once that happens I'll follow up with a more detailed walk through.