Start with simplest problems first. Make a checklist and run down it from top to bottom. Generally it's a good idea to start with the standard network engineering problems before delving into more complex rule tracking. The rule of thumb is to test the easy stuff first!
Start also with the simplest solution first. For example, check for loose or bad cabling, DNS issues, or if the systems being affected are even available before investing a great deal of time debugging the firewall itself. More often than not, problems blamed on the firewall actually have nothing to do with it.
The OSI model, which is explained in the next chapter, is good to use here as a checklist of sorts. You work your way up the OSI model, ruling out dependencies that might be creating symptoms higher up in the OSI model. This further reinforces the process of working on the simplest problems first before moving onto the next complex problem.
