We’ll see what Skype’s official outage explanation is in the morning (they promised to deliver an official explanation), but I have a few theories:
- It doesn’t take three days to restore a botched code patch. Skype is too smart not to be revision-controlling their server code. And these aren’t 500 GB patches, either. They’re probably a few MB in size and can be replicated across Skype’s network in minutes, if not seconds. This “algorithm” explanation makes no sense. If it were bad programming, why didn’t Skype just roll it back from their CVS or their CodeSafe or whatever rev system they’re using?
- A more plausible explanation would be the disappearance and subsequent difficult restoration of centralized authentication data, ie. customer account information. Perhaps somebody forgot to push in a backup tape one night. No joke. This makes more sense to me. Did anybody change a password just prior to the Skype outage and then fail to have that password change reflected after the system came back online?
- The Windows Update theory is out the window. pun intended. Mac users were knocked out too.
