Today’s announcement that Citrix is dropping support for OpenStack has reverberated through the clouderati sphere like a new Justin Bieber song through my niece’s third grade class. Super important but will not matter much when the next idol arrives.
In any case, a lot of smart people have written about it. I’ll leave them to explain the whole thing.
But the post that most caught my attention came from Thorsten at Rightscale‘s. We both share something in common: we both build products that connect to cloud API’s. Including vendor who have API’s that claim to be compatible EC2. This experience, I think provides a useful point of view when thinking about API compatibility. Not to mention it creates a jaundiced view of the human soul.
I’ve said it many times and I’ll repeat it again: it’s the semantics of the resources in the cloud that matter, not the syntax of the API. This means that “API compatibility” has to reach very, very deep to be meaningful. Let me give you a couple of examples around EC2.
Some people say that in the next few years that Infrastructure as a Service cloud deployments will be focused mostly on private clouds. And then they say that enterprises will migrate to public clouds after they have become “experienced” in running a cloud. About a year ago I could really see this story played out. Now, fifteen months after we introduced Cisco Intelligent Automation for Cloud, I have some different points of view. I would have thought that by now that private cloud architectures would have begun to converge to a few standard patterns. This has not happened. The world is still diverging when it comes to both Private and Public cloud architectures.
I do see patterns arising in successful cloud deployments and here are some of the key ones:
#5: Pragmatic Approach: IT shops that come with a long list of RFP requirements and questions take a long time to source a technology provider and to achieve production success. Others that are pragmatic (can I say Agile in their approach) get to cloud quicker and learn from their successes and missteps alike.
#4: They Have a Cloud Instance Roadmap: After a cloud deployment, some IT organizations think that is it, they are done, next project, my move to cloud is complete. Hold it right there, did you know that cloud is not a single step where you through a switch, but a succession of deployments of great scope from one step to the next? A roadmap is needed that covers: hardware, network, storage infrastructure, virtualization technology and release version, management and orchestration software instance version and finally the services that you are offering to the end users and how the service catalog is changing over time. Those that have a roadmap roughed out are generally more successful than those that have a big bang perspective.
#3: Appreciation for Challenge of Management of Change:Moving to cloud is a big change in an operating model; careers are created and new roles are defined. How does an organization move to the new model with different technology, processes and people? When a team proactively manages the change in the non-technical they ensure long term success. It is not just about self service, cloud catalogs, orchestration, domain management and virtualization. It is more about service designers and automation authors and changes in operational processes.
#2: Rise of the Cloud Architect: Since cloud is about a new operating model a new position and role is needed. If you have a cloud project and do not have a cloud architect tying it all together from cost models, to hypervisors, to orchestration and orderable service definitions, you need a organization role tune up ASAP.
#1: A Service Centric Approach: Most people get this one right away. Service centric projects are the key focus for ITaaS. However, I can’t tell you how many times when I am talking to an IT team, the opening bell results in a speeds and feeds conversation around provisioning that piece of infrastructure and that virtualization API. If you ask the question about what services they want to offer their end users for self service ordering you will get a request for more time to answer that question. Service Centric IT shops will take the time to start first with the business requirements and the perspective from the end user point of view. Transform your cloud project approach to a service centric agile project and you will go far.
IT shops deploying clouds over the past year have been focused on Infrastructure as a Service ( http://en.wikipedia.org/wiki/Infrastructure_as_a_service#Infrastructure ) as a way to drive speed in virtual and physical server provisioning, cost savings in operations, proactive service level agreements, and increased control and governance. In one of my blogs I introduced our Cisco Intelligent Automation for Cloud http://blogs.cisco.com/datacenter/the-secret-is-now-out-you-can-simplify-cloud-deployments-with-cisco-unified-management/ and how that addresses both private, hybrid and public clouds IaaS. Key to this is the service catalog and self service portal. Moving to cloud is NOT about taking hundreds of server configuration templates and moving to them immediate self service. All you are doing in that model is automating VM sprawl. They key is defining a limited set of services and options that your end users such as application owners and technical folks can order through a self service portal and manage their life-cycle.
I have talked to many enterprises that want a private cloud; also service providers that want to build a public cloud based upon Cisco UCS / Nexus. They all would like to get to their first generation of cloud very quickly. We have many conversations on infrastructure to use, the right choice of hypervisor, and what use cases to start with. Today I talked to a customer of our Cisco Intelligent Automation for Cloud that wants a differentiating end user portal. Everyone can spin up a VM, but how to do it in a way that will drive adoption and new business, whether it is for the enterprise or service provider? How do we do this at a cost and a timeline that meets the needs of the business and the CIO.
These are not lightweight concerns. I have seen many data center automation projects that focused too much on the provisioning and not enough on the end user experience. The Intelligent Automation Solutions Business Unit at Cisco has been involved with many customers building private and public clouds with our software stack. This includes Cisco IT’s very large private cloud deployment based upon our software. The two services they deployed: CITEIS (Cisco IT Elastic Infrastructure Services) Express and CITEIS VDC (Virtual Data Center) has received broad acclaim. Watch the following videos by Jamie MacQuarrie of the Intelligent Automation team on these two CITEIS Services.
‘Twas the week before Christmas, when all through IT, not a creature was stirring, not even a sysadmin?
Well, not quite. To support the global operations for a Fortune 100 company, the IT staff are always stirring things up at Cisco. But they may be just a little less busy this holiday season. Why? Because Cisco IT deployed a private cloud earlier this year, with a self-service portal and automated provisioning for infrastructure-as-a-service.
This means that employees throughout Cisco can provision and manage the infrastructure resources they need on their own, anytime and anywhere – so our sysadmins can take a break this holiday season (or more likely, they can focus on other IT priorities).