Ask HN: Would you test your web app against simulated infrastructure failure?

Would you pay for a product that provides a third level of testing after unit and integration?

It would allow you to deploy your entire app infrastructure onto a set of lightweight VMs, then let you specify tests (infrastructure failures like packet loss, high latency between nodes; getting pushed into swap; running out of disk; nodes becoming unreachable). Then you could run your application against those tests to ensure it fails gracefully under those conditions.