Syncsort, a provider of big data integration solutions, has announced the availability of MFX ZPCopy, a new software product that can offload mainframe copy processing to zIIP engines, and can be licensed separately as an add-on to Syncsort MFX. MFX ZPCopy is the latest addition to Syncsort’s big data products, which also include DMX-h, a Hadoop ETL solution, that, among other capabilities, provides secure, high performance access to mainframe data from within Hadoop to enable a more complete picture of enterprise data.
“Syncsort MFX runs on nearly 50% of mainframes today, so it is a very big market. The benefits of ZPCopy are that it allows customers to achieve significant operational savings and also run much faster,” said Jorge Lopez, director of product marketing at Syncsort, in an interview.
With increasing copy demand fueled by growing data volumes and rapidly developing big data analytics initiatives, customers are looking for ways to optimize mainframe processing to save money and meet SLAs. “Top financial firms, top retailers, insurance and health care organizations all still rely on the mainframe for most of their operations and when we talk to customers even on the Hadoop front, it is not uncommon to hear that up to 80% of all their corporate data originates on the mainframe,” said Lopez. “They cannot afford to ignore this data - so the more data they generate with the mainframe, the more important it is for them to make this data available to the entire enterprise. This means more data processing on the mainframe like sorting, copying, and moving huge data sets - which creates additional workloads on the mainframe.”
The question for these organizations is how can to make all of this affordable because otherwise they will just keep the status quo and keep doing what they have been doing, which is basically backing up data onto tape “and that is almost the same thing as throwing it into the trash can,” Lopez contends.
After looking at mainframe processing from several customers, Syncsort realized that copy-related processing accounts for hundreds of hours of CPU processing time annually and contributes to batch window bottlenecks, inflating software costs and making it more difficult to meet SLAs. “For some customers copy-related processing can account for 800 hours of CPU time - and on the mainframe that is very easy to make the translation to real dollars. For organizations in that range, ZPCopy can translate into $300,000 of yearly savings,” says Lopez.
In addition, by offloading that processing to the ziiP engines and also doing some optimizations, ZPCopy can also help customers accomplish that processing 25% faster than the original copy function, improving batch window performance, he notes. ZPCopy executes transparently as part of MFX for z/OS, without the need to change JCL or control statements.
“In the end it is not about replacing the mainframe because people who have been in this market a long time know that mainframes are not going anywhere,” says Lopez. “This is about complementing the mainframe and extending the scope and the breadth of all your applications with all that data that is being generated in the mainframe so it is not locked away and it is available to the entire enterprise.”
For more information, visit www.syncsort.com.