Around 10.45 this morning, our supplier maincues started their (announced) power-maintenance. Due to a mistake in our internal procedures, we missed the announcement of this maintenance, which caused our router to go down. Upon our request, maincubes reconnected the power-supply to this router, after which the router came back online.
Even though our network is built redudant, such an interuption can take longer than you would expect. Between the routers, we use the protocl 'BGP'. If one of the routers in a BGP-conversation fails, it may take a few minutes for its peers to detect that. They will be sending traffic to a offline router in the meantime, which will go nowhere because the router is offline. After the timeout of about two minutes, peers will detect that the router is down, and will start taking other routes to the network.
We apologize for the inconvenience this interruption has caused.
