Recently the Shadow team released Shadow 7.2.1. In this podcast Gregg Willhoit shares about the process involved with developing 7.2.1 and some of the exciting features on it. The podcast lasts for 4:22.
You may listen to the podcast here: http://blogs.datadirect.com/media/GreggWillhoit_Shadow7.2.1.mp3
Gregg Willhoit:
Shadow 7.2.1 is an exciting release for us with Progress DataDirect and Shadow, the mainframe product. We’re really excited about 7.2.1 as it gives us a completely new product with regard to mainframe data access, and that is ANSI 92 SQL support for non-mainframe data. I’ll be talking a lot about that at System z. We’re really, really proud of this particular feature. What we did in 7.2.1 for the SQL access is we basically eliminated the need for data replication in many cases. And what I mean by that is by offloading all of the SQL analysis, the data transformation joins, any of the more esoteric SQL functions to the zIIP, as well as all of the network I/O, we’ve basically eliminated the need for data replication in many cases. Shadow in this case has kind of taken what it did in 7.1, the Shadow release 7.1, and applied that entire zIIP offload and enclave SRB strategy to SQL access, and it’s really paying dividends. It actually turns out that complex SQL is even more computationally expensive than XML parsing in most cases, particularly when you’re getting into large amounts of data that is involved in joins. We’re seeing incredible amounts of offload, 99% type offload, so we’re really, really happy.
If you’re interested in SQL access, or even just a zIIP offload story, we have an article coming up in October in zJournal, basically talking about “From SQL to SOAP: The zIIP As a Fungible Virtual Mainframe Appliance,” which basically talks a little more detail about what we’ve done, why we did it, but we truly believe in the mainframe, in System z as the penultimate platform there is for computing, especially business computing, and with the specialty engines, not only the zIIP but of course the IFL and the zAAP, that it presents a very compelling story from a TCO perspective. And heretofore, quite honestly SQL accessed the non-relational mainframe data -- so I’m talking, I’m not talking about DB2 – I’m talking about SQL access to VSAM or SQL to IMS, there are some tools that do this. However, they’re quite expensive to use in most cases.
So what most customers have done is they’ve replicated. They’ve gone through the pain of replicating, offloading that data to a non-z/OS from non-z/OS hardware, running applications under different operating systems with different management and maintenance requirements, and we’ve eliminated the need for that because it’s just as cheap -- actually, I think it’s less expensive to leave that data on the mainframe, and again, use the zIIP as a fungible virtual appliance and provide for ANSI standard SQL access to that critical mainframe data. So we’re excited about 7.2 from the perspective that we truly believe that this brings about a paradigm shift in terms of SQL access on the mainframe and how people will look at it in the future. I think we’re helping at least make a strong case for keeping the data on the mainframe. That’s where it belongs.
In 7.2.1, beyond the outstanding new SQL access that we’ve provided, we’ve done a lot of things to enhance our web services component, particularly in the performance area, and we’re really pleased with what we’re seeing there. We’re seeing much, much greater transaction throughput, as well as less CPU usage. We’ve optimized both the parsing, marshaling/unmarshaling of the inbound and outbound XML and done some interesting things there, and the results -- and the initial customer results are really, really quite impressive. We’re really happy with our improved XML and our zServices support for 7.2.1. And with zEvents, we’ve done some really neat stuff there, particularly with regard to the CICS 4.1 and supporting not only 4.1 but 4.1 and the new events facility within CICS, so that’s really exciting.