Do you understand the risks of deploying on days with limited support?
Loading last updated info...
Deploying on a Friday has traditionally been risky due to limited post-deployment support over the weekend, potentially leaving issues unaddressed. Thanks to DevOps best practices, these concerns are largely mitigated by automated deployments, continuous integration, and real-time monitoring, making the deployment process smoother and more reliable.
Nonetheless, caution is recommended in certain situations.
Figure: Avoid deploying on Friday... unless you have a plan
While it is best to avoid deployments on days with limited support availability, sometimes urgent updates like security patches cannot wait.
How to be prepared?
Pre-Deployment communication
- Inform the team about the deployment and ensure that support staff or relevant team members are available if needed
- Notify all relevant stakeholders, acknowledging that you and the team are prepared to fix issues over the weekend if they occur
Deployment checklist and best practices
Every project should have a document to cover each step of the deployment:
- Pre-Deployment: Ensure that all project components are up-to-date and tested before deploymen
- Staging Deployment: Test the application and check for any errors before moving to production
- Production Deployment: Proceed to production only after confirming no issues in staging
- Post-Deployment: Monitor for any new issues
- Troubleshooting: Outline key contacts for support and provide step-by-step actions to address issues
How to mitigate the risk of a deployment?
- Use Feature Flags: Implement feature flags to control the rollout of new features. This allows for quick disabling of problematic features without a full rollback
- Ensure Comprehensive Monitoring: Make sure all necessary monitoring tools are in place to quickly identify any issues. For instance, you can use Application Insights