Summary： Many think performance problems don’t exist in the cloud, but they do. Deal with them the same way you did before cloud. 許多人以為雲端不存在性能問題，但這些問題確實存在。
Performance issues don't always go away after we migrated our applications and data sets to the cloud. Here’s just a few examples of why.
1.Most performance issues are traceable back to the application and to the hard fact that the applications must be redesigned and rebuilt to deal with the problem.
2.Database response time is a frequent performance issue and often requires a redesign or even an entire change of the database model.
3.Cloud platforms such as Linux and Windows Server often have the same performance limiters as their on-premises versions, including use of memory, user interface management, and external communications.
This list could go on for pages. If you’re sold on the idea that performance must be managed in the cloud, and that the solutions are not typically easy. You need a strategy to approach cloud computing performance management. Below are a few key ideas to get started:
Get down with performance metrics
These systems should help you trend performance behavior over time. They can spot performance issues before they become true issues and link this analysis to self-healing processes, such as changing tunable on platforms based on the changing behavior of applications.
Ops can now take performance issues directly to Dev. This is important because most of the cloud performance issues that a person typically run into are traced back to the application or to the data bound to the application.
However, You can have all of the virtual cloud services in the world on-line running your application and it won’t solve a problem that’s born in programming code.
Click this link for more details： InfoWorld
若喜歡本文，請關注我們的臉書 Please Like our Facebook Page：