Thomas Kurian, President, Product Development for Oracle, talks about the benefits of modernizing and managing infrastructure and applications using Oracle Management Cloud.
We anticipate that 60 percent of IT organizations will move systems management to the cloud by 2020. While it may seem to be a bold prediction, many customers are already well into a cloud journey, while many others have just begun. It's time for a modern, new approach, and organizations know it.
Traditional approaches to systems management weren’t designed for the cloud, others just don't work for it and customers are looking for help. Surveys reveal that less than 10 percent of customers trust their monitoring strategy, which is surprising to most, since monitoring tools have had over 30 years to deliver working solutions. Already over 50 percent recognize that they need a new solution designed for the scale and complexity of the era of Digital Business, Hybrid Cloud, DevOps and Big Data.
A typical company has over 120 different monitoring tools gathering time-series metric and configuration data, and they also generate several terabytes of logs every single day. That means 120 independent data silos of uncorrelated structured data, and tons of relevant (but also un-correlated) unstructured data. Systems Management shouldn't be a massive, custom application requiring IT professionals to spend lots of time trying to make sense of the data; especially as that is time taken away from fixing an issue or prevent a problem from impacting business. Cloud-based management can ingest and process massive amounts of structured and unstructured data and can be put into use by customers on the same day of their subscription. Furthermore, a modern solution requires a machine-learning based approach, not so dependent on human effort or heroics. While traditional "single pane of glass" monitoring tools can gather a good amount of data, they expect humans to go through all the data, diagnose and figure out patterns in behavior and relationships between various events.
With machine learning, algorithms using anomaly detection identify unusual behavior, clustering to identify patterns in data that should be flagged, multivariate correlation to understand relationships between and among data sets and prediction algorithms to reduce the human dependency and help look forward rather than just backwards. A SaaS or cloud-based approach to machine learning delivers instant value, has already been tuned and comes pre-optimized for analyzing vast amounts of data.
Given that most organizations are already unsatisfied with their IT department’s traditional approach or current management tools, and given that a SaaS-based approach offers a low effort and risk option to overcome the major issues with existing approaches, it's not so bold to expect that 60 percent of IT organizations will move systems management to the cloud by 2020.