sbt clean compile test docker
docker push carlpulley/lift-exercise:v0.1.0-SNAPSHOT
docker push carlpulley/lift-profile:v0.1.0-SNAPSHOT
docker push carlpulley/lift-notification:v0.1.0-SNAPSHOT
git clone https://github.com/carlpulley/coreos-vagrant.git
cd coreos-vagrant
INSTANCE=1,2 CLOUD_CONFIG=lift/akka METADATA="lift=true" vagrant up
INSTANCE=3 CLOUD_CONFIG=lift/akk METADATA="lift=true,load-balancer=true" vagrant up
INSTANCE=4 CLOUD_CONFIG=cassandra METADATA="cassandra=true" vagrant up
# You should now have a 4 machine cluster provisioned
ssh-add ~/.vagrant.d/insecure_private_key
vagrant ssh core-01 -- -A
fleetctl start exercise@{1..2} notification@1 profile@1 vulcand cassandra seed
# You should now have:
# - a Lift Akka cluster running with 4 members (2 exercise, 1 notification and 1 profile microservice)
# - a Vulcan load balancer interfacing with the cluster
# - events being persisted to a Cassandra journal/snapshot store!
# TERMINAL 1 - execute the following command:
fleetctl ssh exercise@1
docker logs -f exercise-1 | grep -v akka.cluster.ClusterHeartbeatSender
# TERMINAL 2 - execute the following shell commands:
for i in 1 2 3 4; do
curl -v -H "Host: lift.cakesolutions.net" http://172.17.8.103:8888/exercise/$i
done
# In viewing terminal 1, you should see evidence that:
# - Vulcand has round-robin delivered a GET request to exercise-1
# - cluster sharding is used to route messages to the correct Lift actor endpoint hosting exercise-1
# - evidence of actor spin up and passivation