Architecture Matters - JasonC at Sun: Architecture Matters - A View from the Field

Cloud computing continues to evolve. We are seeing more and more critical apps deployed onto and managed via cloud infrastructures and platforms. A great example is Rightscale's announcement a few weeks ago that they have kicked off more than 500,000 cloud VMs.

Another aspect of the evolution is moving beyond the VM but to grouping servers and services together and treating them as a virtual data center. Sun's Qlayer acquisition in January provided a unique view and model into how these resources can be grouped and managed together. This shifts the lifecycle to managing many if not all aspects of a deployment versus a box or cluster at a time. Typically during a major upgrade, code changes but so does DB tables, presentation tier components, etc. How do you manage the lifecycle of this change vs a component at a time?

NOTE: SDCs = Service Delivery Controller, See the blog below for more info.)

Using VDCs, services can be deployed within a VDC, and eventually even span multiple data centers. A VDC might implement a single business service, very much like Service Delivery Network Architectures' (SDNA) service module (NOTE PDF) concepts detailed here several years ago. As we continue to evolve and release new technologies -- the focus should be on the lifecycle -- and the first step is to realize every data center service out there is transient and has a "time to live (TTL)."

Look for more "evolution" on the SDNA front at Sun's Blueprints program blog.

Preparing to speak with Robert Holt at CommunityOne West in San Fran.

Check it out -- lot's of cloud talks, great keynote from Lew Tucker and Dave Douglas, and of course you might be able to catch our webcast on OpenSolaris DSC (Dynamic Service Containers with Nimsoft's service level management helping us out.) Will post the slides soon.

UPDATE:

Slides posted here.

Statera is putting on a cloud panel in Denver next week, April 30 at the Broomfield Omni. If you get a chance come hear cloud perspectives from Sun, Salesforce, Google, and Microsoft.

Stolen from Statera's event website, here's some of the issues addressed:

COULD YOU BENEFIT FROM ANY OF THESE CLOUD COMPUTING ADVANTAGES?

* Reduced Cost - Cloud technology is paid incrementally, saving organizations money.

* Increased Storage - Organizations can store more data than on private computer systems.

* Highly Automated - No longer do IT personnel need to worry about keeping software up to date.

* Flexibility - Cloud computing offers much more flexibility than past computing methods.

* More Mobility - Employees can access information wherever they are, rather than having to remain at their desks.

* Allows IT to Shift Focus - No longer having to worry about constant server updates and other computing issues, government organizations will be free to concentrate on innovation.

Hope you can come!

As I think about managing resources and services on the cloud (and yes I realize that one persons resource is another's service but...) there's a couple of thoughts I've been sharing with folks and Sun and our customers.

Clouds are a great example of applying "continuous architecture." Continuous architecture is a subject that has a basis in architecture but its really the cross section of biological thinking and the human task of creation. It's the notion of complex adaptive systems applied to something that historically might have been viewed as static (though buildings and houses are anything but.)

In the cloud, static really doesn't apply. A workload might be on one server one minute, on many the next, or on another in the next instant. This may or may not be apparent to the "developer" or service designer, or even the operator of the cloud, in fact its most likely not.

So how is this accomplished? For years we've (via dynamic infrastructure, JxSON, and other projects) thought that workloads are a composite of elements -- to run they require vLANSs, IPs, CPU, Storage, etc. As computer system technology continues to mature, so do the features, especially around the deployment, management, and virtualization technologies that enable much of the cloud today. Anyone that uses GridEngine or other grid/HPC applications -- this is old school. Yeah there's a resource manager, resources are mapped by the manager process, and they run, report back results.

But how does one assemble a bunch of virtualized resources that need to be brought together to meet the needs of a service? Are all the sub-elements assembled via a model and "put on the stack" or do you do this at runtime? How long does this process take? How does a service know when to return its composite resources back to the "pool?" Does it have a default time to live? The model map of the resources need to ensure that they are providing the level of service required by the service.

I was speaking to a customer the other day and every service by default is deployed in their DC with a 30 day TTL (time to live.) You can "purchase" an exception... I won't get into the downstream effects of this thinking -- and their are many but this forethought is significant.

But it illustrates these two processes....the consumption of composite resources by the deployment (run) requests for services.

The Data Center as we Know It is Dead. (Sorry George Gilder)

Officially -- welcome to cloud computing from Sun.

If you haven't seen the news, look at Sun's Cloud Announcement and check it out for yourself. Netnet -- Sun's Cloud business unit is not only building our own public cloud (at SuperNAP today -- but also taking that technology to our service providers, other partners, and our direct customers (private clouds.)

Also, my team, specifically Robert Holt, Mikael Lofstrand, and myself have been working on something called OpenSolaris Dynamic Service Containers (DSC) -- you can join the fun at Kenai.com. This makes it easy to manage containers/zones and scale them up and down across physical nodes -- all via a simple registry.

Additioanlly, we've been working on some IDEs for the cloud -- look at John Kirby's page for some ideas here. This will eventually link to the DSC work above and our own cloud offering based around x86 virtualization (VMIs.) Why is this important??

The deployment model is assembled in the IDE today but the cloud needs to manage the workload. The "model" can't simply live in the IDE, it needs to live in the cloud as well.

Great work team!!