Fuel Module Issues

We have been trying to fix a problem with the fuel module. For some reason, it starts to perform slowly after a while at client sites. Even if Amicus was shut down, the computer would perform slowly until it was restarted.

We performed some automated stress tests on a machine in the office. However, this did not reproduce the problem. No matter how long we ran it for, the problem did not go away. We then tried harder to emulate the client environment. We used the same type of machine, the same printers, everything. This seemed to work. The system started to slow down, memory usage went up, and it would not respond quickly again until after the machine was restarted.

Reproducing the problem was at least something, as we can investigate the issue now. However, we now had to try and find out what was different between the first machine we ran the tests on and the second machine that had the issue.

The strange thing is that we decided to do the test on the second machine again, using the same setup as the first. The point of this was to run the test again, where the only change is the computer itself. I was expecting the system to work correctly, but then it started to fail. For some reason, if the only thing that was changed is the computer itself, the second computer caused the issue, and the first did not. This is annoying, as it will make it a lot harder to find if the problem depends on the computer. Since we get the clients to use the machines that exhibit the problem, this is something we need to fix.

So this week we are going to try to find out why these machines cause this issue.