12 thoughts on “What K-O'ed Skype Last Week”

  1. oh come on, you can’t blame windows for this one. It was a skype problem with some badly written skype code for the windows problem.
    So put down your Apple / Nix soap box and shut up.

      1. And they both are rude.

        aep528: muppets should have clicked the Reply button – IF that was the comment he meant to shut up (rude).

        fwiw: Apple was not referenced in my comment, was it?

  2. No, it’s a Skype problem. They rely on their end users to provide the computing power necessary (instead of running centralized servers), and bill you for the privilege when you want to use premium services. A brilliant business model.

      1. True but their architecture is arguably more resilient and redundant than any major SP (though maybe their software dev/QA/update processes need some tuning).

        Question: do we know if the temporary supernodes were regular, user-owned computers, just promoted, or if they were in fact Skype owned/rented nodes?

  3. Buggy Windows client may be the instigator and forced upgrade could have exacerbated the problem. But lack of some operational procedures are more glaring:
    1. not ensuring that the population of supernodes are diverse (not same OS, not same version of app)
    2. protecting overloaded supernodes from new additions to the network
    3. not preventing new nodes from being added which would increase signalling traffic between the supernodes

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.