Lookery.com

Update: We’re hearing reports of some “HTTP 500 server errors” showing up in place of some ads, though none of us can recreate the problem. We’re in Boston, St. Louis, and Seattle right now, so hopefully it’s a very limited problem. The cause is unexpected DNS propagation error that should decline and vanish overnight.

The good news is that we’re already over two-million ad impressions per day, and Monday is likely our first three-million impression day. The bad news is that we’ve not been as speedy or reliable thusfar as we need to be. The immediate demand for our services surprised us, and we already had to swap infrastructure providers. We expected four hours of downtime to get this done, and unfortunately it took about three times that long. We’ve been communicating with our publishers and marketers privately throughout the process, but email can be a balky thing.

At approximately 1:20pm PDT this afternoon, we went live on Amazon’s EC2 service. Now we can add servers and bandwidth on-demand, which should benefit all parties. If our experience using EC2 at Mashery is any indication, we’ll have much smoother sailing from here than if we used a more standard hosting solution.