The New Year is always a good time for new beginnings, and 2015 is no exception. With the start of 2015, my coverage focus at EMA is shifting and evolving. After 22+ years in the network management space, including over six years as an industry analyst, I am starting a new chapter by shifting my area of research to unified infrastructure management including cloud and on-prem environments.
While it might seem an unusual change, this is actually quite a natural transition. Throughout my years in network management, I’ve always kept an eye on the context of where networks fit into the bigger picture. Networks, after all, are there to serve a purpose – to connect endpoints of various varieties and to deliver applications and services between those endpoints. Networks also represent one of the three legs of the stool when it comes to essential IT infrastructure, together with compute and storage. The three must come together and work together to host and deliver applications to IT end users, customers, and business partners. And while it’s essential to understand how each technology “leg” of the stool is operating, it’s more important to understand how they are all working together (or not) to accomplish the end goal of supporting the organization.
Since arriving at EMA in 2008, my colleagues and I have been watching an interesting and exciting trend: IT operations teams are converging within a steadily growing slice of organizations large and small. By converged operations, we mean a single operations team is assembled that is responsible for all infrastructure monitoring – compute, network, and storage – and quite commonly includes security, application support, and even service management representatives. In a sense, this is war room as a standard operating model. When I first arrived at EMA, we were seeing such changes in about 15-18% of organizations. This percentage grew slowly for years, led largely by organizations that were embracing ITIL and ITSM, but the arrival of cloud caused an inflection point, somewhere around 2010. Now the ramp is steeper, and our last check (early 2014) indicated that nearly 40% of organizations were using some type of converged operations approach. EMA attributes this to the fact that cloud is a service concept, and it simply pushed more organizations down an ITSM-like path out of fear of being obsoleted/outsourced to external cloud providers.
Converging operations is significant for several reasons. First, it causes teams to re-evaluate their management tools, seeking unified solutions that can present monitoring data across all domains together in one place to better support system-wide visibility and converged monitoring practices. Second, it becomes very important to find and agree upon data sources that can be considered objective and independent for use as a primary starting point for monitoring and troubleshooting triage (my favorite is wire data). Third, it’s all about the data – the more the better, although the tools you chose had best be able to keep up with volume, velocity, and variety/variability (hmmm – sounds like big data). And last, everything needs to be built and gathered around a unifying theme, and that theme is end user experience (a.k.a. service quality, or in some cases simply application performance).
Sounds challenging enough, but that’s not where it ends. We also need management tools to converge across virtual as well as physical domains. And further still, we need them to integrate visibility into externally-hosted cloud environments. My favorite here is hybrid cloud, where applications live in part within an on-premises setting (a.k.a. internal private cloud) while other parts live in an externally hosted setting (a.k.a. external private or public cloud), either all the time or part of the time. This means the operations team needs to be able to recognize all parts of the application environment regardless of which type of infrastructure happens to be used, at any and all times, keeping up with change and providing as much visibility as can be gained. And besides the monitoring aspects, there are also the tools for automating and orchestrating configuration and provisioning that will be the new center point of systems, applications, storage, and workload management. Now we’re having fun.
I could go on, and I intend to going forward, but suffice to say for now that there is ample problem space around hybrid cloud management to be poked and prodded. Further, there is a ton of innovation going on within the management vendor community to try and figure out the best way to monitor and manage these mixed, complex, dynamic hybrid environments. I look forward to studying the products that result as they evolve and mature.
Oh, and by the way, since moving over to take this on, I also get to cover a couple of closely related areas that are also evolving with the times, including Data Center Infrastructure Management (DCIM), Workload Automation (WLA), Data Center Automation (DCA), and Converged Infrastructure management.
Just in case anyone is concerned (two, maybe three of you) I’m not leaving network management completely behind. Tracy Corbo and our newest analyst, Shamus McGillicuddy, remain on my research team and will lead network management coverage for EMA. They will keep me well grounded and in tune with what is happening in networking and network management. And with them covering my prior area, I’ll be free to stick my head in the clouds and do some looking around.