How does agile development improve mainframe performance? (1)

During the Second World War, there was a classic joke about conflicting military intelligence-the two opposite things could not be combined. The same effect can be achieved if "business" is used to replace "military. Some people think that it is impossible to combine cutting-edge and expensive agile development processes with old-fashioned and clumsy large hosts, especially when system administrators try to manage the performance of legacy large host applications.

Don't worry! The tools and user best practices developed by the vendor prove that this is completely feasible. Before we get excited, let's take a look at our position.

Agile development is unstoppable

I have written many articles about agile development. Here, I think there is no need to repeat the highly positive analysis content. In short, almost all software developers are talking about agile development. Scrum, agile Business Intelligence (BI), and even the continuous integration of tested modules and large-scale New Version execution have all made steady progress in various software development institutions.

In the past four years, agile developers have learned how to adjust by number of people and code length. At the same time, vendor tools are managed from the past application lifecycle (ALM) unit development is more related to agile developer processes. These tools also expand their scope. Therefore, agile application lifecycle management now not only means uninterrupted interaction between testing and coding, it also serves as a coordination function between developers and operations.

Unshakable targets of large hosts

Also in the past four years-especially in the past two years-senior management has discovered the obvious blind spot for system administrators over the years: application performance, not only the normal running time of the application, it is important!

For a long time, IT only concentrates on keeping applications running. The complexity of today's architecture and the difficulty of adjusting legacy applications make us have to face up to the root of this problem. At present, there is no good application response time, performance problems and initial test interruptions. Good results only result in customer dissatisfaction and low employee efficiency, otherwise, the slow running time is prolonged, and the trial and error fixes are interrupted. Interruptions and slow operation will make key applications unavailable for a long time. This will have an impact on the sales and production cycles and even lead to a stop. Sometimes it can cause serious losses to the Organization.

This issue is particularly important for large host data centers and their internal legacy applications. Legacy applications not only help commercial operation, but also become a more and more important place for Linux applications. If the response speed of Linux applications is as slow as that of snail crawling, it will have an important and profound impact on global online users. Let alone the Business Intelligence Application-there is no conflict here.

The increasing attention of enterprises on legacy applications makes the application performance management (APM) tool components finally become the target of enterprise IT investment. At the same time, manufacturers are also investing in the "arms race" to develop APM tools for new architecture combinations. Obvious options include IBM Tivoli, CA, and Gomez/Compuware. Unknown options include Precise Software, which adds a balance between the analysis content, and recommended restoration methods. I strongly recommend using this tool in the overall architecture. The tool was introduced as a result of significant progress in these tools targeting key application performance management requirements of system administrators. The result was monitoring: understanding all software levels, supervise applications that span the cloud and other networks, master the effect of load competition and compound applications, use accumulated knowledge to perform faster and deeper root cause analysis, and propose better repair suggestions.