Alebra specializes in moving and integrating data across applications running on z/OS and other operating systems, which drives innovation on the IBM Z platform by our customers. One such customer is a large international bank that aims to quickly meet the increasingly complex needs of both its retail and commercial customers. They place a high priority on innovation and are early adopters of solutions requiring innovative software and hardware that delivers value.
Like many IBM Z customers, the bank is in possession of a large amount of data residing in DB2 and IMS databases. These databases are unmatched in delivering fast and reliable service to the bank’s core online transaction and batch processing systems, but much additional value could be extracted from this data by exposing it to “big data” analytical tools running on Linux on Z. Accomplishing this task proved to be a challenge; data was unloaded to flat files and copied to a data lake, reading the data off disk with BSAM and moving it via TCP. This was a slow, resource intensive process, consuming excess CPU time and disk space. Because of this inefficiency, the bank was unable to realize the value of analyzing this data in real time.
Alebra’s Parallel Data mover was the solution to this problem. The PDM Subsystem Interface was used to send the data straight from the output of the unload utilities into Linux on Z with no interim stops on disk, moving over high speed HiperSocket connections. This is accomplished almost entirely in a zIIP, maximizing speed and minimizing CPU consumption on the z/OS host. Data in transit was compressed with zEDC, further cutting down elapsed time.
“Thanks to IBM Z features and Alebra’s Parallel Data Mover, we are able to keep our crucial business data on z/OS, taking advantage of the platform’s strengths of reliability and speed, while still making that data available to Linux applications at next to zero impact to our batch and online transaction processing operations.”
---Mainframe Solutions Strategist – CFO’s office