While I am not Nate Silver to make predictions about what is going to happen in 2013, this is my wishlist for 2013. Please note that none of these wishes is completely new, but none is prime either. If you provide a prime solution in any of these areas, I would be happy to be correct. Please leave your comments.
Hotly debated topic – from why do we need this to how to do this. While there are attempts to define what Interoperability really means, it is also one of the mostly malleable terms. A favorite argument used both for and against OpenStack – depending on how you spin it, someone even claimed 2013 will be the year of InterOp. My sincere hope is that the industry arrives at a closure on what InterOp really means. If any one platform can provide that by end of this year, that will be great!
Cloud Security seems to be an after-thought; relying hypervisor security at the best. While compliance has been looked, how one can reliably say that cloud (platform) is secure? What tools/ methodologies do we have? Do we have threat models of cloud management platforms? This might seem to be so 2010 (as in people when people would ask is cloud secure), but I am looking for a quantitative methodology to say that cloud is secure. Know of any? I am not entirely sure of what Cloud Security Alliance does, but would love to learn more without having to pay any membership fees.
3) Standard Benchmarks
My cloud is faster than yours! Each provider has one’s own definition of resource units. How to quantitatively compare providers or even platforms? How to quantify a platform’s performance? While I am aware of proprietary/ in-house tools to measure, I haven’t come across one like SpecInt or TPC. It is not a OS until it can be measured.
4) Better Disaster Recovery mechanisms
Where to start? Lot has been talked about the outages last year – AWS Christmas eve outage or its postmortem, or Azure outages. However, it is not entirely prudent to leave it up to cloud providers. One has to design for failure, and I hope to see more better practices/ patterns around that.
5) Usage Optimizations
Cloud resource usage optimizations – this has enough jargons to throw in a pitch statement for an aspiring startup, wait a minute. I already know of one – MetricsHub, a TechStars – Azure Accelerator 2012 graduate. There is a lot of monitoring/ reporting tools available – both from the providers(Amazon CloudWatch), or as services (Cloudability, NewVem, Cumulux(Opstera)) or cleanup tools (JanitorMonkey), but I am looking at something that can optimize based on the usage patterns. Easier said than done!