Wrapping up our seasonal look at 2013 predictions, I thought I’d go a little higher up the stack to see what’s happening in the middleware and application layers and how this might affect infrastructure.
One of the first things that strikes me is the extent to which automation will increasingly act as the arbiter of relationships between applications and infrastructure. Automation has been a facet of data environments for some time, but only now are we seeing the IT equivalent of the auto-pilot, in which automation stacks can not only identify problems like traffic congestion or failed resources, but then reroute data to safer terrain and even effect repairs if the issue is on the virtual layer or higher.
About the only time an actual tech has to get involved is when hardware needs to be replaced, and even that could soon seem quaint if trends such as robotic hardware management, which is currently being developed by modular data center designers, takes hold.
Since the cloud is more about services than infrastructure, it’s not surprising that application performance is emerging as a key component of widely distributed architectures. And from the looks of it, application performance management (APM) is set to make some significant gains next year. According to Compuware, we should be on the lookout for tools that increase visibility across increasingly disparate internal and external infrastructure, including new mobile platforms. Also, expect to see a greater emphasis on application lifecycle management that extends from development, through testing and into production.
In many ways, applications themselves are starting to incorporate many of the management tasks that IT once dominated. In business intelligence, for example, Forrester notes a distinct trend toward more self-service capabilities that allow users greater control over functions like ad hoc querying and relationship mapping. This transition is unlikely to result in a complete shift toward user management, but will send most enterprises on a quest to hit the “sweet spot” between optimal productivity and effective governance.
This need for increased self-management is likely to increase as many larger, complicated applications move to the service model. As IDC’s Matt Oostveen noted to Australian clients recently, Big Data and business analytics platforms are simply becoming too large to be cost-effectively supported on internal infrastructure, pushing most SMBs and a good number of mid-level enterprises to seek cloud-based solutions. And with on-premises infrastructure on a steady path to increased convergence, data intensive apps not only run the risk of bogging down their own data environments, but everyone else’s too.
Indeed, the trend toward services will probably transcend top analytics and operational management applications rather quickly toward more general-purpose platforms like SharePoint, says b2bmarketing.com’s Chris Boorman. This will require a high degree of intelligent collaboration across data, application, cloud and infrastructure environments, which in turn will impose new requirements on master data management (MDM) and data integration platforms. Again, the top challenge will be to maintain a significant level of centralized governance without hampering business and data agility.
As I’ve mentioned before, this is all part of the general breakdown of what once were pretty rigid spheres of influence among management platforms. Infrastructure management used to concern itself with the basic functioning of hardware. Application management and middleware confined themselves to software environments and systems. But now that everything is virtual, these tools must consistently reach outside their traditional comfort zones to embrace the world around them.
In the end, it’s not about managing software or platforms or infrastructure. It’s about managing data.