DCIM — Updates And Trends

Only a few months since I authored Forrester’s "Market Overview: Data Center Infrastructure Management Solutions," significant changes merit some additional commentary.

Vendor Drama

The major vendor drama of the “season” is the continued evolution of Schneider and Emerson’s DCIM product rollout. Since Schneider’s worldwide analyst conference in Paris last week, we now have pretty good visibility into both major vendors' strategy and products. In a nutshell, we have two very large players, both with large installed bases of data center customers, and both selling a vision of an integrated modular DCIM framework. More importantly it appears that both vendors can deliver on this promise. That is the good news. The bad news is that their offerings are highly overlapped, and for most potential customers the choice will be a difficult one. My working theory is that whoever has the largest footprint of equipment will have an advantage, and that a lot depends on the relative execution of their field marketing and sales organizations as both companies rush to turn 1000s of salespeople and partners loose on the world with these products. This will be a classic market share play, with the smart strategy being to sacrifice margin for market share, since DCIM solutions have a high probability of pulling through services, and usually involve some annuity revenue stream from support and update fees.

How Big Is The Market?

I’ve received a number of inquiries about the state of the market, and I will admit that my answers have been pretty fuzzy because the larger companies are not prone to part with details of their revenues, and the small privately held firms are equally reticent. Through a series of interviews with customers and triangulating revenues of the firms through their competitors' lenses, with an attempt to factor in the various biases, I think that total revenue for DCIM solutions, not including services, is in the neighborhood of $200 million. Currently my best educated guess is that Emerson (counting all its former piece part products as well) is currently number one in revenues, with nlyte and Schneider competing for number two, possibly with a slight edge to nlyte, with the three of them accounting for perhaps $100 million in revenue.

This is a highly arguable estimate, and if anyone cares to contribute additional data, I’m all ears.

Product Directions

The major changes I’ve seen in the last few months are an increasing trend toward further segmentation of product offerings, with multiple low-end entry options as well as versions targeted toward service providers, with, as you would expect, granular multi-tenancy capabilities and flexible options for reporting and billing.

While all vendors are not currently at the same place in their implementations, all of them have their eyes on two emerging requirements: better visibility and management of the server workloads, primarily through VMware and Hyper-V integrations, and more granular cost and resource accounting.

Major System Vendors — Missing In Action

HP and IBM (despite IBM having a basic DCIM capability in its Maximo product) have not been active participants in the market to date, although they have all implemented what amounts to DCIM implementations as part of data center services engagements. Aside from having substantial IP distributed across many unrelated products, HP is beginning to lobby for its top-down data center management architecture, but that is a long haul to reach the market. My best guess is that there are additional cycles of acquisition by both HP and IBM, possibly Dell to build out their product portfolio.

Emerging fringes for functionality include early discussions of embedding sophisticated reliability modeling capability to assess risks for business services running in the data center.

Looking Forward

I am anticipating rapid growth in this solution category, possibly in excess of 100% PA as Emerson and Schneider add a large number of feet on the street and users increasingly learn the value of these solutions.

I’d be interested in hearing about your DCIM plans.

Comments

DCIM as an integration layer

Rich, thanks for this post.

Let's try answering a simple question - "Why do we need data centers?" - Data centers are needed to enable the delivery of efficient IT services related to the business needs of an organization, with appropriate levels of availability, security, scalability and reliability – and, increasingly, operational efficiency and cost effectiveness. One of the primary requirements of DCIM is to establish a much needed bridge between the IT or logical side of the data center and the facilities or physical side, thereby achieving that primary goal of enabling the infrastructure to be used to effectively and efficiently support those IT services and satisfy business needs. As of now, there are various point solutions available to monitor or manage particular subset of functions, processes, infrastructure or type of logical entity within the data center. But a big challenge is that bridging the gaps between these disconnected or loosely connected systems, where attempted, has traditionally required a good amount of manual data consolidation, ad hoc analysis, challenges in reporting and limitations alerting and control, leading to a lack of transparency across systems and lots of inefficiencies.

Since data centers are critical but complex environments, DCIM works best when you view and manage them in a joined-up way, regardless of which specific infrastructure components and systems are present. In reality power distribution, cooling, BMS, servers, virtualization software, IT management software, and so on, come from different manufacturers. So it’s the ability of a DCIM system to work across these effectively – without vendor bias – which is really important. Also, it’s true that basic DCIM functions are starting to become better recognized and more widely available across some products, but the real innovation in DCIM is emerging across boundaries, at the intersections between these classes of system, achieving greater convergence and optimization in resources, power, cooling, space, IT services, virtualization, cloud, business processes, and service levels.

Editor's note: Commenter is a VP at CA and a DCIM subject matter expert.

complicated, missing key features

DCIM products out there seem to be way too complicated to administer and to run. Many not so advanced features require add-on's or small tools which appear to be an after-thought or look like (and work like) somebody has programmed them in a day or two because a customer asked for the capability. Major version changes and consolidations do not help.

Given the multiple parties involved in the management and use of a data center (as the previous commenter alluded to) requires processes and work flows. Yet very few products have solid and easily customized workflow built-in, with role based security, to appeal to these various groups.

We looked last year, decided to wait and are just now starting to look again for a DCIM. Maybe we'll just stick with our homegrown database and combination of spreadsheets and Sharepoint lists for another year.