Short hosting interruption Sunday, November 25

On Sunday, November 25, 2012, at around 1.46PM Central European Time, a problem occurred in the primary load balancer of the hosting platform. Our automated monitoring system initiated an automatic fail over to our secondary load balancer at 1.50PM. The fail over took a few minutes and was complete at 1.53PM.

At 1.56PM, the monitoring software noticed that one of the OpenX installation on our platform was not working correctly running from the secondary load balancer, and sent out an alert to the duty engineer. While he was working on the secondary load balancer, another engineer had already begun restarting the primary load balancer. Normally, the switch back to the primary load balancer would be scheduled for a slow time during the night, but in this case it was decided to switch back immediately because of the problem with the secondary load balancer.

At 2.23PM, the switch back to the primary load balancer was initiated, and after 3 minutes at 2.26PM Central European Time, our platform was fully operational again.

We apologize for the inconvenience caused by this incident, which resulted in two 3 minute time frames of unresponsiveness within about 30 minutes of each other.

Share this on:
  • Twitter
  • Facebook
  • LinkedIn
  • email