Deployment and Release Management
Manage the integration, test, and deployment of solutions, and their delivery into production.
Improvement Planning
Practices-Outcomes-Metrics (POM)
Representative POMs are described for Deployment and Release Management at each level of maturity.
- 2Basic
- Practices
- Adopt some automation in deployment.
- Use a centralized build server.
- Outcome
- Manual processes may still be in effect for certain deployment types.
- Metric
- % of deployments that are automated.
- Practice
- Put a designated team in place to manage releases.
- Outcome
- There is a team to manage and provide consistency of release into the production environment.
- Metric
- Existence of designated team to manage releases.
- Practice
- Implement procedures and controls for high-profile releases into production.
- Outcome
- There is an agreed approach for releases into the production environment.
- Metric
- Existence of procedures and controls for high-profile releases into production.
- 3Intermediate
- Practices
- Implement a pipeline, automating quality assurance steps and progressing builds across test environments.
- Ensure that deployment-led practices are integral to the activities of delivery teams.
- Outcomes
- There is a deployment process in place.
- There may be a two-tier deployment process, with legacy applications still requiring manual steps.
- Metrics
- % of deployment steps that are automated.
- # of solutions delayed pending approval.
- # of roll-backs.
- # of successful releases.
- Practice
- Put a standardized release management approach in place, with validated release procedures and templates.
- Outcome
- Solutions are released consistently and on time, with minimum disruption or negative impact.
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.
- Practice
- Put in place a process for consistently evaluating the potential impact and risks of releases, involving a range of stakeholders.
- Outcome
- Stakeholders are aware of upcoming releases and their impact.
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.
- 4Advanced
- Practices
- Ensure that deployment processes are analysed and improved as part of deployment design.
- Continue to refine and update automated processes and supporting tools on an ongoing basis.
- Outcomes
- The deployment process is fully automated.
- Multiple platform types are supported.
- Metrics
- % of deployment steps that are automated.
- # of solutions delayed pending approval.
- # of roll-backs.
- # of successful releases.
- Practice
- Put in place a standardized release management approach across the board.
- Outcome
- Solutions are released consistently and on time, with no disruption or negative impact.
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.
- Practices
- Put in place a proactive approach to impact and risk monitoring.
- Implement automated scheduling, allowing for out-of-hours releases for the majority of changes.
- Outcome
- Stakeholders can access information about releases at all times in a number of different ways.
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.
- 5Optimized
- Practice
- Ensure that full repeatability of automated deployment and release processes is in place across all platforms and test environments, and into production.
- Outcome
- Roll-back of deployments and releases is robust and repeatable.
- Metrics
- % of deployment steps that are automated.
- # of solutions delayed pending approval.
- # of roll-backs.
- # of successful releases.
- Practice
- Put in place a clear, well-defined, flexible process for managing the release of solutions into production.
- Outcomes
- Releases do not involve service or performance degradation (e.g.
- Canary releases or Green/Blue deployment approaches in place).
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.
- Practice
- Implement automated scheduling, allowing for out-of-hours releases as needed.
- Outcome
- Releases do not involve service or performance degradation (e.g. Canary releases or Green/Blue deployment approaches in place).
- Metrics
- % of successful releases.
- % of stakeholders aware of upcoming releases and impacts.
- # of solutions delayed pending approval.