Microsoft’s WGA failure attributed to "human failure"
By Jonathan Schlaffer
You may remember this past weekend that many users had trouble validating their legally obtained copies of Windows Vista and XP. The company is now willing to talk about it to a limited extent and has said the problem was due to a human error.
It’s like the infamous computer of the Discovery said, “things like this have cropped up before and they have always been attributed to human error.” Besides being a deranged homicidal computer, there is truth in that.
The truth of the matter is that the servers were functioning properly and the checks and balances would have worked if preproduction code had not been sent to the server. Once the code was sent the whole shebang went down.
Microsoft is not saying the servers went down, because they remained functioning but were functioning incorrectly. The good news is that if the servers were down, the installations would not have started pushing up daisies (died for those who’ve never heard that before). If the servers are still working but not functioning correctly, you will be identified as a pirate but if the servers are simply down then nothing happens, which also makes sense if you think about it.
Once the dust settled and according to ArsTechnica less than 12,000 computers were affected by this problem, not a big deal in the millions upon millions upon millions of computers that run Windows. What’s worrisome is that there was no automatic system to say the servers were not functioning correctly and redundancy (if you can call it that) was present in only the most basic form. Technicians rolled back the servers to the previous code but the effects of the WGA problem were not so easily undone.
The system is currently being upgraded with more redundancy so this (should) not happen again. Google has lots of redundancy, billions of people rely on Google, more than Microsoft (in terms of numbers), anyone remember the last time Google went down? Didn’t think so. Maybe Google will let Microsoft take a look at its systems.
On another note, higher ups at Microsoft are wondering why users were told it could take 72 hours (or more) before the issue was fixed.
Related:
Entry was posted
on Wednesday, August 29th, 2007 at 5:46 pm
under
Read more entries by Jonathan Schlaffer.
Stumble It!
