Forget about mapping your toolkits' domain to that of your monitoring platform, we already did that for you. All the built-in dashboards are designed to take your toolkits' domain into account and present data to you in a familiar and intuitive way.
First class support for Akka, Play Framework and all the asyncrhonous and non-blocking goodness that comes with them, right out of the box. See your actors, routers, dispatchers and HTTP endpoints behavior like never before.
See how garbage collection affects the behavior of your application, keep track of hiccups and get access to the data you need to keep them under control. Long stop-the-world pauses will never go unnoticed again.
Get support from professionals dedicated to deliver value using the same toolkits that you use. Email, chat, video conference, in-house trainings, you name it and we will find a way to smooth your integration and monitoring experience.
All the metrics collection and runtime instrumentation is provided through Kamon, a open source tool for monitoring JVM applications with years of production usage, a constantly growing and welcoming community and the smallest possible performance footprint.
Don't like then instrumentation? You can change it. Want to integrate a different toolkit? Get your hands dirty and write the integration yourself. As long as you follow the metric naming conventions your data will show up just fine in our dashboard and you will have entire control of what runs in your servers.
On your environment, Kamino is just another backend for Kamon, you can use it side by side with any of the existent backends for other open source and commercial monitoring alternatives and go back if you are not satisfied (although, we will always work hard to keep you with us!). Selecting your monitoring platform couldn't be any easier.
When data needs to be aggregated for display or storage, all the raw data is aggregated, always producing the real percentiles and metrics on any time period that you select. No summaries, just reality.
Choose whatever percentiles make sense to you and your users, get alerted when anything goes beyond the acceptable limits. You are not only allowed to set the percentiles that matter, you can change them any time, having all the raw data allows us to calculate any percentile on the fly.
Configure any percentile limits for any metric you want. Get alerted if instances stop reporting unexpectedly or if your service levels are breached. You get to tweak how and when the alerts are triggerd to avoid false positives and annoying unnecessary alerts.