Woodworking is one of my passions and I often use wooden pallets as my source material. Regardless of what I build—whether a shelf, chair, or bookcase—I always use the same things: Wood, tools, and a plan that shows dimensions and steps to put all the bits together.
The other day, it struck me how similar this is to how organizations digitally transform and become data-driven. Rather than a fancy Adirondack chair, your plans focus on growing, connecting, or protecting your business. Your raw material is data instead of wood, and you use multidisciplinary analytics rather than woodworking tools like a hammer, saw, and sander. Just like the tools, your analytic functions need to work in concert although the results are insights rather than sanded rough cuts left by the saw. If not, your observations may be misguided and business value not fully realized, if achieved at all.
Multidisciplinary analytics are the tools, the different workloads you need: data engineering, data warehousing, data science, and operational analytics. To achieve your business goal or solve for your particular use case, you need the full complement and in varying ratios throughout the project.
Working on the same shared data, the workloads also need to integrate on another level: data context. Uniquely, Cloudera’s machine learning and analytics platform have a fundamental characteristic called the Shared Data Experience (SDX) that provides just that.
When transient cloud infrastructures are used to complement existing on-premises investments, establishing and capturing this data context is essential for success. We addressed many of these requirements when we introduced SDX a year ago and extended its capabilities to Cloudera Altus.
There are several permutations when it comes to the infrastructure and storage that you can use to run these diverse analytic workloads, ranging from on-premises to public (IaaS or PaaS) and private cloud or any hybrid combination. Therefore, it’s crucial that we provide complete consistency across the entire realm of data context: security, governance, lifecycle, control and data catalog.
Rather than wait until everything everywhere is fully implemented, we bring out the support in stages. IaaS and private cloud are already supported through our reference architecture. With the release of SDX for Altus workloads as-a-service, we’re now supporting the second most common combination: sharing data and metadata between customers’ own Cloudera workloads deployed to the public cloud (IaaS) with Altus Director and those managed in the public cloud by Cloudera as a service (Altus PaaS).
Organizations are no longer constrained in choosing how and where they deploy their workloads based on the ability they have to manually manage the data context between them. Any combination of workloads can now take advantage of the most appropriate infrastructure. The benefits are realized as soon as more than one workload is deployed.
Here are a few real customer examples:
- A retailer needs insights on web-based sales. IT has tight control and is running its highly customized Cloudera Data Warehouse workload 24×7 as an Altus Director-deployed Cloudera Enterprise cluster. The company also has a transient Altus Data Engineering workload to bring the data into the Data Warehouse environment. This has the benefit of being highly repeatable and automated as well as providing the lowest TCO. With SDX, the data context from the transient workloads is directly shared with the Data Warehouse, providing critical auditing information and ensuring trusted data is always used for analysis.
- A large financial services organization is required to provide always-on analytics to their customers and doesn’t want to depend on a single cloud vendor for that. Using a team of just three administrators (rather than the 15 they needed before) and thanks to SDX providing single consistent context, they can now deploy their Cloudera analytics applications to no fewer than three public clouds in less than an hour. In addition to their operational cost savings, they went from concept to production in just six months and achieved substantial business benefits.
In all cases, SDX helps IT reduce its management overhead and costs, as well as lower risk for the company as a whole. The business further benefits from improved governance, greater agility, and faster business insight. With SDX now on Altus, you have maximum cloud flexibility. You can run clusters anywhere and manage them in any way, with the unified control IT needs and the self-service business and data professionals demand.
For me, the joy of woodworking is discovering new techniques and solutions, like glueless joints. I start by trying them on a small scale before using them in anger; it helps me familiarize myself with the approach and the tools for the job. You can do the same for your data-driven projects: Add SDX to your Cloudera Enterprise deployment and follow the steps in our reference architecture documentation. With our free 30-day Altus trial, you can try the shared data experience for yourself in your environment.
With that in place, what will be your first project?