- High single instance reliability. I have ~1000 EC2 instances; every week I get a notification how an instance has been restarted to move it off failing hardwa

- High single instance reliability. I have ~1000 EC2 instances; every week I get a notification how an instance has been restarted to move it off failing hardware. I never got these in my DC - Consistent large scale workloads at low latency. Nothing beats low latency then bare-metal servers. Also, it costs significantly cheaper running large clusters on bare-metal then in the cloud. If you move to the cloud, you need to use the cloud services else your costs are astronomically higher. - If you already have datacenters that have been paid off why move to the cloud. Use your own data centers and run a hybrid cloud. I think there is a reason allot of big tech companies roll their own cloud or do some kind of hybrid approach. Whe nyou running a app that is core to your infrastructure you do not want to rely on another persons hardware you want 100% control of everything. I have had route 53 cause issues, S3 cause issues, EC2 servers get stuck and there is nothing you can do but wait 15 min for the enterprise SLA of AWS. In a business critical system, 15 min is not good enough. You need your operations techie to be sprinting to the rack to hit the reset button. Just my experience.