When most IT professionals think about Big Data, their minds naturally turn to Hadoop. But for many IT service providers, the most immediate opportunity for building a managed service may have little to do with Hadoop. Instead, massive amounts of machine data is now being natively captured in a repository developed by Splunk specifically optimized to provide operational intelligence to IT organizations.
IT organizations tend to favor Splunk to capture that data because they don’t have to go to the trouble of transforming it into another format to analyze it. That eliminates the need for an Extract, Transform, and Load (ETL) process that is both time-consuming and expensive to manage.
The challenge is that many IT organizations have internal Splunk expertise. For that very reason, Accenture has become one of the first IT service providers to create a managed service for operational intelligence based on Splunk. Accenture also revealed recently that Hawaiian Telcom has become one of the first customers of that service.
In the future, Accenture also plans to incorporate Splunk into a security information event management (SIEM) As-a-Service offering as well.
Other MSPs that have built a practice around Splunk include BT Counterpane, Cloudmark, T-Systems (a division of Deutsche Telekom), and Verizon Business.
Benefits for managed service providers
Of course, MSPs can make use of Splunk to optimize their own IT operations at scale. Most MSPs in actuality are a conglomeration of IT functions. The cost of delivering those services winds up being aggregated across multiple customers to keep expenses under control. The challenge that creates is a massive amount of IT complexity across multiple classes of application workloads. Making use of Splunk to provide operational IT intelligence in real time helps discover issues before they become a serious problem, which minimizes any potential disruption in IT service.
Just as importantly, Splunk can help put an end to the IT blame game. Instead of wasting hours trying to figure out what went wrong and who is responsible for it within the IT organization, MSPs can readily identify the root cause of the issue at hand. Right now, the vast majority MSPs are spending days trying to identify a problem that only takes them a few minutes to fix once it's discovered.
Before too long Splunk will also serve as a resource for the development of machine learning algorithms that will enable MSPs to operate at even higher levels of scale by relying on increased automation. Of course, no IT organization can successfully automate processes without visibility into them. Unless they understand all the dependencies that exist between the applications and the IT infrastructure that makes up the IT ecosystem they are trying to manage, it’s only a matter of time before something will go wrong. The difference between something going wrong with a manual process and something going wrong with a process that's automated, however, can be nothing less than profound, especially when it comes to unraveling all the potential damage inflicted on a unprecedented level of scale.
None of this means that MSPs should not embrace IT automation, but it does mean that MSPs need to proceed with care.