第二十章 云架构 - 学到的教训
If you are completely successful in navigating the complex and competing priorities levied on every cloud computing solution, your efforts will fail unless a successful implementation follows. Although implementation is outside the scope of this solution design text, we would like to share with you the lessons we have learned in these early years of cloud computing architecting:
- To be a successful cloud solution architect, you need to obtain and maintainexecutive sponsorship. Make sure governance control points are built into the transition process (an example control point: ensure financial controls on pay-as- you-go elastic compute model do not result in runaway costs!).
- Analysis of an organization's application portfolio as a whole is key to efficiency and the breadth of value delivered by transitioning to a cloud platform. Management oversight and review processes for legacy application transitions to IaaS platforms should be modified to reflect their software-only nature.
- Most customers are only aware of a few large cloud service providers (for example, AWS, Azure, Google, Salesforce, and IBM). They also may be limited to selecting a single CSP platform (that is, C2S). This does not reduce or eliminate the need to evaluate the economics and performance aspects of a transition strategy to the broader marketplace.
- Lack of IT standards or a failure to enforce those standards results in differences between your development, test, and production environments. This significantly reduces your ability to leverage automated testing tools and delays your cloud transition. Developers must be educated on Application Performance Monitoring (APM) capabilities, service management/monitoring capabilities, web and mobile analytics, and alerting and notification solutions.
- The most severe challenge when adopting cloud computing is around cultural change. A focused and dedicated informational and educational campaign should be in place to support this type of transition. A lack of cloud computing education and understanding is one of the most significant organizational risks.