Consisting of two 320 node clouds operated by Indiana University and the Texas Advanced Computing Center at University of Texas Austin, Jetstream is a first of kind system funded by the National Science Foundation. Jetstream was conceived four years ago, and early operations started two and a half years ago. In this time there have been a number of mistakes made as well as pitfalls avoided in both the construction and operation of this distributed cloud resource. Causes of major outages, design choices, pain points, as well as comparisons to traditional HPC systems will be covered.
