I kind of "met my Waterloo" today afternoon in fixing the production Crystal reporting database using server cloning. It was merely a month ago I proudly figured out a way to get things done effectively and efficiently, but ever since then with a haughty mind I had believed that a task like that would be nothing more than a piece of cake — and that was exactly the mentality of mine early today when I was starting to work on the production server, the last step to wrap up the project.
And unfortunately I did not document what the magics I did last time to get the things done in less than a hour.
For almost 4 hours, because I missed a thing here and there, the database was not even accessible, but an unnecessary listener was mysteriously created instead. Everything looked so bad and at one point I thought the server had all been messed up and a re-cloning was the only sad choice left for me.
Fortunately, the QA server is available for me to use as a reference to dig out part by part what I had done the last time, and meanwhile through the listener log I found out what had gone wrong. It was not until the end of the day that I managed to get the business back on track!
Lesson learned now: the first thing I did right after the database up and running was to create a brief document to have all the steps included. I know for sure that I have no desire whatsoever to go through the same pain again.