z Systems Solutions Consultant
In July 2017, IBM announced the IBM Open Data Analytics for z/OS product, the next evolution of the z/OS Platform for Apache Spark that was announced in 2016. IBM expanded its support to include Anaconda and Python environments, thereby serving a larger community of data scientists.
With the IBM and Rocket Software collaboration, this product continues to enable access to mainframe data such as DB2, DL/I, ADABAS, IDMS, and other data types. Using its easy-to-use Data Server Studio tool, queries quickly enable access to many types of mainframe data, and federate to distributed data, such as Oracle and SQL Server.
After hearing about this new product, your IT organization may be curious to find out more. Will this new product provide the value it touts? Implementing new products takes time and money. So, how does a company justify the use of this new tool?
Business Case for IBM Open Data Analytics for z/OS
There are many benefits to the analytics community by implementing this product. Let’s explore the financial benefits that can justify a Proof of Technology project.
There are two areas of cost saving to consider – ETL tool and the ETL processing costs. Most enterprises can identify the software maintenance costs, as well as their staffing costs, in support of these tools. But, many companies do not effectively capture the costs of the ETL process at the various data sources and loads. These processing costs can be significant, especially when running on mainframe General Purpose processors. As the mainframe capacity increases, so does the software cost in support of that capacity. IBM Open Data Analytics for z/OS executes on specialty processors (zIIP), not Mainframe General Purpose processors. These specialty processors do not affect mainframe software pricing models. Therefore, the ETL tool maintenance costs and the mainframe software costs, in support of the ETL process, become cost savings.
Data Access Requirements:
The analytics industry continues to evolve with new tools, such as Jupyter notebooks, Tableau and others. These are the tools of the data scientists. The business should provide a quantifiable cost benefit of utilizing today’s tools versus using outdated tools purchased years prior.
Machine Learning has a growing interest in the data analytics industry. When companies are considering this functionality, they may be considering a new hardware platform in support of it. The new IBM z14 is designed to enable the performance needed for Machine Learning through such technology as SIMD and large in-memory support. IBM Open Data Analytics for z/OS is a pre-req for Machine Learning on z/OS. This implementation on Z Systems is a cost avoidance, from a purchase of a new hardware platform, to enable Machine Learning capabilities.
Data scientists want access to real-time data for their decision process. The business is challenged to quantify the lost revenues due to incorrect assumptions on day or sometimes week-old data from a data warehouse, data mart or data lake. As an example, today’s campaigns require access to real-time data for more nibble reactions and informative decisions, in an effort to capture Generation X or Millennials market share. What’s the financial impact of a misinformed campaign?
Querying actual source data has led to the concept of data lakes. Here, raw data is stored so tools can quickly be tweaked to access the raw data formats. Now, take this concept one step farther, and eliminate the need to replicate the data into the data lake, but instead, query the data at its source. The savings here is cost avoidance of the storage equipment and processing of the data lake’s replicated data.
The data lake could house data from many data types, not only sequential files and Db2, but IMS DL/I, ADABAS and other old mainframe data sources. The data lake data repository could become complex to query. IBM Open Data Analytics for z/OS includes the built-in capability to mask the data type from the query the data scientists execute. Here, there is cost avoidance to writing another ETL program or access vehicle to old mainframe data types. In addition, this masking benefits the business by enabling the joining of these old data types in a federated query. The business needs to quantify this benefit.
Each time data is replicated to another data source, the data is exposed to missing security rules and hacking. By keeping the data on the mainframe, the query leverages the mainframe data’s original security specifications. Also, the mainframe has rock-solid encryption, using security capabilities such as True Random Number generator for keys, encrypted paging, and tamper protection appliance with Secure Service Containers that protect the data, both at rest and in flight. What’s the cost of failed data security on your company’s data?
Now that we better understand the business case for IBM Open Data Analytics for z/OS, the next BLOG will explore the steps to the planning and execution of a Proof of Technology (POT) for your company. IBM, Rocket Software and Mainline offer many resources to help you with your POT.
Please contact your Mainline Account Executive directly, or click here to contact us with any questions.