COMP598 - McGill

Simple Elastic Cloud with Load Balancer

A basic cloud management system that will manage and monitor a cluster of machines and use them to run jobs. Nodes and Pods are depicted by docker containers and networks, respectively.

Group 3:

Alice Kang, Hana Zhang, Joshua Roccamo

---

Dependencies:

• docker
• haproxy
• nginx
• ...

---

Files to start when starting the cloud:

On comp598-user@winter2023-comp598-group03-02:

• resource_manager.py
• resource_monitor.py
• cloud_dashboard.py
• elastic_manager.py
• request_load.py <- for testing different loads on the elastic manager

On comp598-user@winter2023-comp598-group03-01:

• cloud_toolset.py

On comp598-user@winter2023-comp598-group03-03:

• light_proxy.py
• medium_proxy.py
• heavy_proxy.py

---

After running the above files:

• use the cloud_toolset commands to run the toolset (node/pod management) and monitoring commands
• access the cloud dashboard by opening this link in a web browser: http://127.0.0.1:3000/cloudmonitor/dashboard
• to see the stats live without going to the dashboard: watch 'echo "show stat" | sudo socat stdio /run/haproxy/admin.sock | cut -d "," -f 1-2,18,47-48,59-61,74,77 | column -s, -t'

---

Supported Toolset Commands

cloud init
cloud pod register {pod_name}
cloud pod rm {pod_name}
cloud register {node_name} {pod_id}
cloud rm {node_name} {pod_id}
cloud launch {pod_id}
cloud resume {pod_id}
cloud pause {pod_id}
cloud elasticity lower_threshold {pod_name} {value}
cloud elasticity upper_threshold {pod_name} {value}
cloud elasticity enable {pod_name} {lower_size} {upper_size}
cloud elasticity disable {pod_name}

---

Supported Monitoring Commands

cloud pod ls
cloud node ls {pod_id}
cloud log request
cloud log pod {pod_name}

---

Supported End User Requests

curl http://10.140.17.108:5001/light/app
curl http://10.140.17.108:5002/medium/app
curl http://10.140.17.108:5003/heavy/app