turbine-ec2

Turbine Instance Discovery based on EC2 tags.

Why

At ContaAzul, we have used Netflix/Hystrix for circuit-breakers and Netflix/Turbine to aggregate Hystrix streams.

Turbine has Eureka InstanceDiscovery, but we don't use it, nor plan to use it in near future.

It seemed simple enough to just use our already defined tags to wire services to Turbine.

That's what this project does.

How it works

You will need to configure the config.properties file just as before. The difference is that you can specify which tag and value should be looked up to find instances of each cluster.

Example:

turbine.aggregator.clusterConfig=my-svc,other-svc
turbine.instanceUrlSuffix=:8080/hystrix.stream

# generic tag, used whenever a specific tag is not specified
turbine.ec2.tag=SERVICE

# specific tag and tag value for `my-svc`
turbine.ec2.tag.my-svc=SVC
turbine.ec2.value.my-svc=svc-value

# needed to access AWS api
turbine.ec2.aws.access=AWS_ACCESS_KEY
turbine.ec2.aws.secret=AWS_SECRET_KEY
turbine.ec2.aws.region=REGION

If you set a custom tag value via turbine.ec2.value.${cluster}, the value of this property will be used as filter. Otherwise, the cluster name will be used.

If you set a custom tag name via turbine.ec2.tag.${cluster}, the given value will be used as tag:${tag_you_provided}. Otherwise, the value of turbine.ec2.tag will be used instead. If none is set, an exception will be thrown.

That's it.

Deploying

You can either download the turbine-ec2-core.jar and wire it with your own code, or download the turbine-ec2-web.war and deploy it in a servlet container.