Contingency plans
When deploying new versions on the production things can go awry.
This is a scenario report in case the production goes down, what will happen and what steps need to happen to get it back up as soon as possible.
- Notify DevOps engineers to start working on finding root cause (check resource usage, bottlenecks, etc)
- Check logs - error logs, access logs for suspicious activity
- Disable plugins one by one and check if the issue persists
- Change the theme to the default one (if possible using WP-CLI)
- Flush cache and clear transients
- Reset permalink settings
- Check logs - error logs (max 5 min)
- Add maintenance page notice
- Revert the latest stable release
- The whole process shouldn't take more than 20 minutes in total if possible, to minimize the impact to the end users.