The first step is knowing if there are slow applications in the architecture. How are the applications communicating? Connections and ports in use, and what are the hard- and software components.
Most of the times if I ask a customer about the architecture of the application, they have no idea. Is this the fault of a administrator? No, the problem is in the history of ICT, the way we work today is in silo’s, while our transactions are going over different layers. You need a helicopter overview across the complete environment.
Yes, but how can I do this? One of the tools we use is Lansweeper or TADDM from IBM, these are excellent tools for discovery.
It gives you an overview of how applications and systems are connected. Another method is Dynatrace Ruxit, it will show you incoming and outgoing connections.
After we have discovered the landscape of the architecture we start focusing on the end2end transactions and infrastructure.
Yes we don’t focus on infrastructure first, that’s the old method; start with the application discovery and find the slowest component in the chain. It’s like playing the weakest link.
Continued in part 2 soon
24 februari 2017