Day 2 Challenge: Your primary compute is going down for 4 hours, 7 days from now
“Shit happens but life goes on.” Anonymous
Challenge Statement
Hopefully you have split your hosting service away from your database service. Which ever X service you have utilize for Day 1, which for my case would be web app. Let’s assume that it is going to be out of service, and you are getting notified about it 7 days prior to it.
While it does not happen that often, lets assume this is beyond our control, and lean towards this worse-case scenario. The service X will be unavailable 7 days from now in all region due to a planned maintenance, and you a strict >15 mins SLA to meet.
Prerequisite and Requirements
- Understanding of stateless and stateful application
- Location of source code associated with your CMS system
- Code repository services, devops knowledge and optionally FTP knowledge
- Possibility of code version difference (bad in this case) between different compute services running identical services during rollout.
- Mitigation strategies to point 4
- Load Balancer and Sticky Session
- Warm vs Hot site
Thought Process
To be completed