A project to setup Mesos/Marathon Cluster along with Hydra in automated way

This project helps you setting up mesos-marathon cluster along with [Hydra] (https://github.com/lake-lerna/hydra) on cloud (GCE, AWS, Azure). For now, only GCE is supported.
To setup a cluster with Hydra, use following commands

1. Do Gcloud configurations
   -> curl https://sdk.cloud.google.com | bash
   -> exec -l $SHELL
   -> gcloud auth login
   -> gcloud config set project <gcloud project name>
   -> gcloud config set compute/zone <zone-name e.g us-central1-f>
   -> gcloud config set compute/region <region-name e.g us-central1>
   -> gcloud config set component_manager/fixed_sdk_version 0.9.81
   -> gcloud components -q update
   

2. Clone and Install hydra-deploy
   -> git clone https://github.com/lake-lerna/HydroSphere
   -> pushd HydroSphere
   -> pyb install_dependencies
   -> pyb install
   -> popd
   

3. Setup Mesos-Marathon Cluster and setup Hydra on master node
   First of all, modify the setup configuration file. A sample file is given in hydra_deploy with name setup_config.ini .
   -> pushd HydroSphere/src/main/python/hydro_sphere
   -> python hydro_sphere.py --config_file setup_config.ini --deployment_id <unique id for deployment>
   --ssh_key_file <path to your publick key>
   

You can access mesos on masterip:5050 and marathon on masterip:8080

4. Login to master node to run test cases
   List current instances
   -> gcloud compute instances list | grep <user-name>
   Please note down the IP of your master node. It would be with name <emailid-deploymentid-*>
   -> ssh <master-ip>
   -> cd hydra-master
   -> source ../venv/bin/activate
   -> hydra zmq
   

TO setup only mesos cluster without hydra, you can give following command
-> pushd HydroSphere/src/main/python/hydro_sphere
-> python mesos_marathon_setup.py --config_file setup_config.ini --deployment_id --ssh_key_file <path to your public key>

If you want to setup hydra later on Mesos/Maratho cluster, you can issue the command
-> python hydra_setup_script.py --deployment_id <unique id for deployment> --instance_user <instance_user_name mentioned in your config file>