IT Ops monitoring: It's time to move beyond the single pane of glass
The "single pane of glass" approach to IT operations monitoring and management no longer makes sense for many IT organizations. You must enable diverse users to access data using any tool that suits them. Forcing everyone to adopt a generic platform to interact with the data created by numerous IT monitoring tools doesn't work.
Instead, enable all your colleagues and business stakeholders to access operations data through a lens that makes sense for their individual goals and responsibilities. After all, the technology exists to cost-effectively share IT Ops data across your organization without overwhelming your systems.
The pain of the single pane
The single pane of glass (SPOG) was, and still is, the desire of many IT operations management departments that wanted to do the right thing by encouraging all users to adopt a uniform solution.
Unfortunately, many have found that this approach falls apart when you try to force people with diverse goals to use a SPOG within a single “uber-tool.” For the last decade, many of the widely used tools have waged war over whose platform would serve as the single pane, with their customers having little valuable to show for it.
Each tool wanted all the data flowing into its platform and all eyes glued to its UI, regardless of why the data was being tracked.
The multi-pane of glass approach
IT operations professionals have been looking at this problem from the wrong angle. Getting everyone to use the same tool is as likely to be successful as trying to get everyone in a group to agree that iPhone is better than Android, or vice versa.
Things have already moved far in the other direction—toward diversity in the way data is consumed. Homogeneity is great for server and central compute resources because it allows easy management and automation. But for the devices or software that consume ever-growing streams of data, multiple purpose-built channels are best for supporting diverse, best-of-breed ecosystems—not to mention optimizing the ROI from those sources. This multi-pane of glass (MPOG) paradigm gives organizations the flexibility to accomplish that.
BYOD and the rise of sexier UIs for applications have already begun pulling users away from SPOG solutions. The genie is already out of the bottle. You can't go back to telling everyone that they have to use the same devices or software and conform to “our IT standard."
The best-of-breed approach, using the right tool to address a specific challenge, is going to beat an all-in-one solution any day.
An analogy: Email data and multiple modes of usage
Consider that the email format hasn't really changed since IETF RFC 2045 (launched over 20 years ago, in 1996), but think about how many apps you use to consume email. Personally, I now have three apps that I use regularly, because no single one meets all of my needs:
- For searching through email, I use Gmail.
- For CRM integration, I use the Salesforce inbox.
- For seamless calendar integration, I go to Outlook.
Each app displays the same emails, but with specific features and functionality that draw me to use it, depending on my goals.
Now, apply this analogy to more complicated enterprise IT tools, and the problem becomes more challenging. Each user has a unique set of requirements that a specific tool fulfills best. This isn't to say that a SPOG isn't a laudable goal. But only execs will use the SPOG on a regular basis. As for application support teams or other first-responder operations engineers, they’ll gravitate toward the specialized, domain-specific tools when an issue occurs.
Sure, you can add more functionality to an application performance management tool to make it more capable of identifying network issues or database issues, but when it comes to getting the members of the Oracle team engaged, they are going to use Oracle Enterprise Manager. Period. End of story.
Don’t lock into a single repository or data lake
Consider a world where the domain-specific tools support a streaming consumption model, whereby data from third-party tools can be ingested and displayed within the UI. We'd then have the level of cross-domain, end-to-end visibility that the SPOG was designed to address.
The first tool that opened up this field was Splunk, which has now become the default integration approach for much of enterprise IT. But there are many other analytics and visualization tools that end users want to use. Plus, no single IT operations management tool is a one-size-fits-all solution.
Embrace multiple consumption models
The time is right for accepting multiple visualization methods that map to the unique needs of your users. Consider the idea that consumers of IT operations data want to consume it in different ways.
Some only want events of a specific type or severity. For others, only data related to the impact on end-user satisfaction matters. Management wants dashboards—showing all green, of course!
Ultimately, if you can move toward an approach that enables domain-specific tools to ingest data from other sources, you are more likely to achieve the goal that has seemed mostly unattainable: immediate, situational awareness and the ability to resolve incidents quickly and efficiently. You can do this without changing the tools your IT teams want to use on a day-to-day basis. In order to quickly and easily implement MPOG, you need integration middleware that collects data from your disparate tools (the data sources) and lets you seamlessly distribute this data to the appropriate destination—mediating and normalizing events, metrics, and topology as necessary.
All parties get to see the data in their own way and act on it accordingly. And as a side benefit, some of the silos and walls between teams can come crashing down. It’s time to move beyond the constraints of the single pane of glass and open up to a multi-pane reality.