Skip to content

Comment on What do you think was the cause of the M$ Outage? Will we ever know?

You are viewing one comment and its replies. View all comments

Comments

  • Infrastructure of this size is just code. It’ll spin up instances, route traffic and balance nicely.

    They probably added a trivial timeout to their health check. Say, if an instance doesn’t respond within 60 seconds it gets cycled out.

    Probably worked fine on Monday morning when they rolled it out.

    It probably took them until Saturday night to remember that it takes two minutes for an instance to launch, so every instance will be cycled out before it tags in.

    Just a wild guess about the kind of miss that might’ve caused it. I haven’t read it heard anything.

    5

Sign in to comment. Sign in