The Chaos Monkey
The Chaos Monkey is a service running on the Amazon Web Services which is looking for Auto Scaling Groups (ASG) and terminates instances (virtual machines) for each group randomly. The software is designed flexible enough that it works well on the platforms of other cloud providers. The service is fully configurable, but by default runs on ordinary weekdays from 09.00 until 15.00 o’clock. In most cases, Netflix has written their applications so that they continue to work when an instance has some problems. In special cases, this does not happen consciously, so that their own people have to resolve the problem in order to learn from it. The Chaos Monkey thus runs only a few hours a day so that the developers are not 100% rely on him.
Find more information about the Chaos Monkey and the Simian Army under Netflix: Der Chaos Monkey und die Simian Army – Das Vorbild für eine gute Cloud Systemarchitektur.
More services from Netflix Monkey portfolio
- Denominator: A tool for managing multiple DNS provider.
- Odin: An orchestration API that can be called by Jenkins and Asgard, the Netflix deployment tool. It is designed to help developers deploy more convenient.
- Recipes: These are blueprints, which can be used to roll out several Netflix components together more easy.
- Launcher:Rolls out the final blueprints by pressing a button.
- Mehr Monkeys: Other Netflix Monkeys are to follow later this year, including the Conformity Monkey, the Latency Monkey and the Latency Howler Monkey. The Conformity Monkey ensures that all relevant instances are placed equivalent. The Latency Monkey simulates latencies and more errors. The Howler Monkey latency monitor whether a workload meets AWS possible limitations and reports it.
The Chaos Gorilla which randomly simulates the crash of an entire AWS Availability Zone and the Chaos Kong that simulative shoots an entire AWS region should also follow soon.
Although Netflix would offer its “Monkeys” across multiple clouds, so that e.g. also OpenStack users can refer to it. However, with the steady release of its HA test suite Netflix plays more and more in Eucalyptus hands.
Eucalyptus Cloud lets you build a private cloud based on the basic functions of the Amazon cloud infrastructure. What is at Amazon for example an AWS Availiablty Zone represents in Eucalyptus a “Cluster”. Thus the Netflix tools allow similar HA functionality testing in the private / hybrid cloud, like Netflix already uses it themselves in the Amazon public cloud. In addition, Eucalyptus will certainly integrate the Netflix tools in its own cloud solution in the midterm to have an own HA test in their portfolio.
The release of the Netflix tools under the open source license will strengthen, not least, the cooperation of Amazon Web Services and Eucalyptus but rather make Eucalyptus more attractive for Amazon as a takeover target.