It is an odd fact of nature that sometimes actions in a certain direction really do produce opposite and equal reactions. In a blog on the EMA website titled Contrarian Accelerators I looked at the cloud in terms of actually accelerating service management adoptions. There are lots of reasons for this, not least among them the increased pressures on accountability, cross-domain visibility and control, and the need for IT to adopt a clear position of leadership in optimizing services for business value across a wider range of options.
As you’ll see in the blog, I really call cloud a “perverse accelerator,” though I picked “contrarian” for reasons of political correctness just in case cloud defamation groups should ever emerge within the general politic. But perverse is the word I prefer because cloud is such a mix of technologies, and has generated such (largely unhelpful) hype that the actual value of this maelstrom of enablers too often gets presented as a kind of religious cult, i.e., the journey to the cloud.
This is the fourth in my series of columns on cloud adoption. In this column, I’m going to focus on the core management technologies to go from just cloud to EMA’s notion of the Responsible Cloud. With a nod to ITIL v3 let’s look at cloud adoption from a lifecycle perspective:
Service strategy and service design
Service strategy requires planning and optimizing services based on business value and impact. Cloud makes this dialog more essential than ever because it poses so many new options, including the ever popular (or still popular) notion of circumventing IT altogether for constituency-specific SaaS solutions in particular.
The technologies we look at below can support this dialog, but they can’t supplant it. This process centers in communication, process and politics and many of the prior columns on cloud are targeted at supporting this requirement.
Service design, i.e., application development, is often served by cloud; platform-as-a-service (PaaS) in particular. It’s arguably the only truly relevant constituency (Who else is really eager for just platforms to work from?). But application developers are also beginning to look at how API choices can impact application modularity in Web and Web 2.0 environments. Bundling applications with infrastructure to provision services in a cloud-like manner (e.g., CA’s acquisition of 3-Tera) is just another technology to watch. But no one to my knowledge has fully worked through the complexities of how SOA, Web 2.0 and virtualized infrastructures will all come together 10 years from now.
Awareness and planning and service transition
Okay, “awareness and planning” is not ITIL v3 (or v2 for that matter). But it is really a necessary preliminary to service transition in cloud or hybrid environments. And here is what you should look for:
Good discovery technologies, including application discovery and dependency mapping – Of course, this applies 100 percent to internal cloud. But it’s also key if you’re looking to plan where and how to use outside service providers or hosted resources (hosted/private). It follows the common sense idea that if you don’t know what you’ve got, you can’t begin to optimize it. And this would include cross-domain infrastructures, infrastructure-to-application interdependencies, and application-to-application interdependencies, as well.
Cloud increasingly puts pressures on real-time or “run-time” (capturing changes when they happen) awareness. EMA just completed a Radar of seven leading app dependency vendors: ASG, HP, IBM, ManageEngine, OpTier, Service-now.com, VMware — some more focused on performance-centric discovery, others more on classic ADDM. Check it out if you want to learn more about what these vendors have to offer.
CMDB/CMS – Strong insights into interdependencies need to be leveraged in a more systemic approach. This area deserves a column in itself but the short answer is that there is a linear and positive equation between effective CMDB/CMS deployments and effective cloud computing assimilation. But cloud is also pushing CMDB design towards more versatile, federated and real-time design points. Think of this as a system of reconciliation based on the power of service modeling rather than a monolithic physical database. EMA is currently doing a radar with more than 20 vendors not including the “big four:” BMC, CA, HP and IBM.
Automation – Effective automation for provisioning cloud services is another column but it basically breaks out into a number of key pieces all of which should ideally be able to leverage common insights from the CMS described above:
- Configuration management: systems, network, storage, etc.
- Having cohesive policies to configure physical-to-physical, physical-to-virtual, and virtual-to-virtual is essential.
- Workflow for service desk-process control.
- Service catalogs that can support self provisioning as well as effectively publishing information about cloud services so that consumers understand what their options are.
- Cloud-unique provisioning capabilities (e.g., bundling applications with infrastructure as described above, or VM specific options that need to be assimilated into larger processes).
- Run-book automation to unify processes across all of the above, along with automated diagnostics when problems occur. EMA uses the term IT Process Automation here.
Monitoring and User Experience Management – Cloud is collapsing performance management, change management and capacity planning into a single thought. Ultimately, you will want to achieve a reconciled system of modeling in which user experience, configuration, capacity options and service interdependencies can all be viewed cohesively.
User Experience, by the way, is getting an added boost from cloud; including service provider-related services. (Most cloud providers do not have adequate SLAs and have no clue regarding user experience. But I have seen them knuckle under when key clients insist on leveraging their own monitoring tools.) Two leaders in user experience management are Compuware/Gomez and Keynote, although there are many other vendors you should also consider here.
Continual service improvement
Accounting/metering/infrastructure utilization/demand profiling – Understanding how services are being used in terms of customer priorities, infrastructure impact, and costs is key to the effective assimilation of cloud services. These insights can close the loop back into service strategy as critical information in planning how to prioritize new services (cloud-related or otherwise). Almost every vendor defines accounting according to its convenience, but truly granular insights into how and why services are used often remain elusive. IBM’s Tivoli Usage and Accounting Manager is one of the more robust. But watch this space for innovation.
Capacity optimization in service context – For the last decade, this area has been more voodoo than virtual, let alone real. But in part thanks to improving technology and analytics, and in part thanks to the “perverse acceleration” from cloud, this is beginning to change. BMC and CA have most recently made announcements in this area, and HP has strong analytics for financial optimization in service context. Apptio and n(i)2 are also vendors to watch for innovation here.
That’s about it … for now.