ENTERPRISE MANAGEMENT ASSOCIATES®
Analyst Corner
December 2011 EMA Analyst's Corner:
CMDB/CMS: The Move to Federation and Optimizing Cloud
On February 2, I'll be delivering a Webinar on "Optimizing Cloud for Service Delivery: Where, How and Why Infrastructure and Services are Coming Together in Real Cloud Deployments." Then, on February 9, I'll be delivering a Webinar on an "EMA CMDB/CMS Use-Case Radar: From Database to Federation." Both will be driven from research projects (Optimizing Cloud from a global respondent base of more than 150 cloud adopters that's underway as I write this). The EMA Radar Report involved many months assessing and reassign offerings from ASG, Axios Systems, BMC, CA Technologies, IBM, iET Solutions, HP, LANDesk, and ServiceNow.
I know there are many in the industry who would consider promoting both cloud adoption and CMDB/CMS adoption as somewhat contradictory pursuits. Cloud is the umbrella term for a media wave that's become an industry wave and is, in many respects, the darling of the times. Cloud symbolizes what's new, fast, cost-effective, and very quick time to value (in many respects with some justification—although not in all respects). CMDBs are, conversely, associated to many with a non-dynamic vision that resulted in failed promises, lengthy times to value or often no clear value at all. In at least one vendor presentation I saw CMDBs framed as the failed, "traditional approach" to service management and service delivery.
The truth is, however, that CMDBs have never been traditional, especially as they merge into federated Configuration Management Systems (CMSs)—which is just now beginning to happen. CMDBs represent an approach to service management that is at least as revolutionary as cloud and maybe even more so. After all, we've had Software-as-a-Service solutions for many years now, and virtualization—at some level or other—goes back to the mid-90s with VLANs and even before with mainframe technologies. Moreover, moving information and investments in and out of the data center (aka cloud) may turn out to be not quite as interesting as decoupling management investments into trusted sources defined in terms of most relevant data (including performance KPIs) linked by a service model that can associate physical and logical interdependencies with increasingly higher levels of automation.
The problem with CMDBs in their initial vision was quite possibly that the vision itself was too revolutionary. Couple this with marketing hype, overeager salesmen, and industry analysts with minds like cardboard boxes (which is after all how markets are created) and you have turned Aphrodite into the Bride of Frankenstein in a few short years. Cloud has been a lot more fortunate, in part, because it's harder to place false boundaries around it, and the industry commitment to innovation around cloud has accelerated to extraordinary degree. Cloud is—to use the words of one of my analyst peers more than 15 years ago when SNMP was overtaking SNA—"where the money is."
What should be pointed out as well is that—believe it or not—CMDB/CMS and Cloud are becoming increasingly symbiotic. In last year's "Operationalizing Cloud," for instance, EMA did a cross analysis that suggested that those IT organizations with CMDBs in place were:
- 1.4 times more likely to reduce complexity of management via Cloud
- 1.4 times more likely to improve service resilience via Cloud
- 1.6 times more likely to accelerate deployment of existing services via Cloud
- 1.6 times more likely to accelerate the creation of new services via Cloud
- 1.5 times more likely to increase infrastructure flexibility via Cloud
- 1.7 times more likely to expand revenue channels via Cloud
And why is this you might ask? Was it another case of bias overcoming objectivity in scrutinizing data?
Well, I'll admit that I was happy to see the results. And I'll also admit that we saw parallel values for Application Dependency and Discovery Mapping (ADDM), which by the way was included as relevant in the CMDB/CMS Radar. There was also some similarity as well when it came to "service management dashboards with advanced analytics," and "IT process automation or runbook," albeit with a somewhat different menu of results with the exception of ADDM, which was virtually identical to the CMDB in achieved Cloud benefits.
As CMDB deployments become more federated, more dynamic, and more CMS-like (and here Cloud serves as a positive catalyst), they are also becoming increasingly able to meet the demands of Cloud. Both CMDB/CMS and Cloud deployments face some common challenges: the need for a stronger cross-domain orientation than exists today in many IT organizations, the need for selective visibility into areas of critical relevance, and the need to support multiple stakeholders—often with very different perspectives so they can quickly come together for purposes of cohesive decision making. When service providers get involved via Cloud, the need to associate logical and physical interdependencies and define process "owners" becomes even more paramount (unless IT organizations are willing to cast their lot with the legal defense fund for self-protecting SLAs and their associated attorneys.).
I can't promise that I can examine all of the ramifications of the CMDB/Cloud intersection in my upcoming Webinars, but I can promise you that I'll do my best to capture the highlights of what I know will be a lot to think about between now and early February when I get to share the results with you.
