Friday, May 23, 2014

Back to the future: Why z/OS mainframe is the ideal cloud platform

Back to the future: Why z/OS mainframe is the ideal cloud platform


By Pat Toole and Frank DeGilio
Fifty years ago, IBM CEO Thomas Watson Jr. unveiled the System/360—the first general purpose computing system specifically designed for business—calling it “the beginning of a new generation, not only of computers, but of their applications in business, science and government.”
It was this vision for computing that made possible the creation of bar codes, ATMs, electronic stock trading, online travel reservations, weather modeling and countless other inventions that have changed the way the world works.
Watson’s statement is as timely and true today as it was then, in the context of the evolution that’s occurring within IT via cloud computing. The System/360 ushered in the era of flexibility, adaptability, and economy for IT. It pioneered the concepts of backward and forward compatibility, the ability to add capacity as you needed it, and protecting the client’s investments in their technology, applications and data.
As we talk about the next inevitable phases of cloud, like platform as a service (PaaS) and software as a service (SaaS), we encounter some very specific themes and challenges that cry out for systems with those qualities:
• How do I provide multi-tenancy with the greatest number of people sharing the same machine as possible, securely and efficiently?
• How do I dynamically scale an environment immediately without waiting to provision?
• How do I provide each user a unique performance experience on the smallest number of systems possible?
• How do I keep track of how much capacity each user is using?
• How do I support thousands of users with a handful of IT professionals?
If there were an ideal operating system that could provide multi-tenancy, dynamic scaling, a unique user experience, fine-grained usage monitoring, and that could support thousands of users with a small group of admins, PaaS and SaaS solutions would be simple. Oh wait, there is—z/OS.
What was once only accessible to COBOL and Assembler programmers is now available to anyone. Today, a team of young IBM hardware and software developers in Poughkeepsie is creating simple web-based interfaces for z/OS, which will allow any programmer with web development experience to use this powerful platform for PaaS and SaaS.
IBM also continues to refine the technology and the delivery of mainframe computing for cloud applications. The new IBM Enterprise Cloud System can support up to 6,000 virtual machines in a single system, provide a secure multi-tenant environment and dynamically share resources across enterprise workloads, with higher system efficiency and greater scalability that lowers the total cost of Linux cloud deployments by up to 55 percent over comparable x86-based cloud infrastructure.
And the new “IBM MSP Utility Pricing for System z” pricing model, delivered through IBM Global Financing, provides consumption-based pricing designed especially to make mainframe technologies more widely accessible to Managed Service Providers. This consumption-based approach allows MSPs to focus on building their businesses, rather than on the cost of their infrastructure.
These innovations unlock the potential for z/OS and the modern mainframe to be the cloud platform of choice, enabling people to create business solutions built on the most secure, available, reliable cloud infrastructure ever known.


http://thoughtsoncloud.com/2014/05/back-future-zos-mainframe-ideal-cloud-platform/

Sunday, May 18, 2014

Make the Extraordinary possible: IBM Mainframe50

Make the Extraordinary possible: IBM Mainframe50

possible
IBM Mainframe50
The mainframe is a revolutionary computing system that is transforming business, whole industries, and the world as we know it.
By continually adapting to trends and evolving IT, we’re driving new approaches to cloud, analytics, security and mobile computing to help tackle challenges never before thought possible. We do that by collaborating with our clients to build better technology every day.
But the pioneering innovations of the mainframe all serve one enduring mission: Deliver game-changing technology that makes the extraordinary possible and improves the way the world works.
See the link below to watch a replay of the April 8 NYC Mainframe50 event.

The Engines of Progress

They transform our lives. Their visionary endeavors make our world smarter and more connected.
Previous

First National Bank

Critical financial services to Africa’s unbanked 80 percent
FNB's "bank in a box" enables millions of individuals across Africa to startup new businesses, save for college and improve lives
 

Met Office

Vital reports to empower industries and save lives
The Met Office generates thousands of weather forecasts a day to help global industries and governments make crucial decisions
 

Walmart

Helping 250 million people a week save money and live better
From a Five and Dime in Bentonville, Arkansas to a retail superforce serving customers in 27 countries around the world, Walmart has built an empire on one simple value: the customer always comes first.
 

First National Bank

Critical financial services to Africa’s unbanked 80 percent
FNB's "bank in a box" enables millions of individuals across Africa to startup new businesses, save for college and improve lives
         

 

GENERATION z

They will create the future.
Every year, students at over 1,000 universities across 67 countries are learning to solve tomorrow’s grand challenges with the mainframe.
The top 43 students from around the globe met in New York City for the first ever Master the Mainframe World Championship.

50 extraordinary years

They invented the modern world. Organizations and leaders have reinvented business, transformed industries and helped solve some of the world’s greatest challenges with the world’s most innovative and trusted platform. These achievements are woven into the fabric of industry and society.

Sabre reservation system

The world’s first automated passenger reservation system, Sabre, is the basis for today’s web-based travel industry. Launched on two IBM 7090 mainframes, the application written in the 60s is able to process billions of transactions a week today.

Wednesday, March 5, 2014

Enterprise COBOL 5.1—Where Tradition Meets Innovation





COBOL is the most ubiquitous programming language for developing business applications. Despite that its specification was created more than 50 years ago, COBOL is still running the world’s most critical business applications. Many of us might not realize how much we rely on COBOL applications in our day-to-day lives. Every time we use an ATM, book an airline ticket, process a check or make an insurance claim, we’re using a COBOL application to process the transaction, and there’s a good chance this application is running on System z.

Modernizing Enterprise COBOL

Enterprise COBOL is a premier enterprise-class COBOL compiler on System z, powering many business-critical applications. Time proven and reliable, COBOL has a long history of delivering new and innovative features for developers to modernize key applications, and increase productivity.
Modern System z servers—z10, zEnterprise 196 and zEnterprise EC12 (zEC12), for example—are designed with high-performance microprocessors and are reshaping the IT landscape. COBOL developers are expected to deliver new functions faster with higher quality and performance. As the hardware become more complex, it’s becoming increasingly difficult for application developers to simultaneously handle business logic and performance issues in the applications they’re working on. They need state-of-the-art compilation technology to provide advanced optimization and hardware exploitation to deliver leading-edge performance so they can focus on the business logic. To satisfy this requirement, Enterprise COBOL needs to transition to a new technology base to ensure timely delivery of System z hardware exploitation and advanced optimizations.

Improving Performance

Announced in April, Enterprise COBOL for z/OS V5.1 was re-architected to include a new optimization framework based on proven technology shipped in Java on System z. IBM has been shipping this technology since 2006. Unlike Java, this new framework for COBOL is static in nature. It provides advanced COBOL specific optimizations, and full z/Architecture exploitation (z900 through zEC12). A new option, ARCH, enables users to specify which z/Architecture to optimize their applications for. Increasing performance of COBOL applications provides savings in CPU utilization and the ability to meet service-level agreements.
Enterprise COBOL V5.1 runs on z/OS V1.13 and later. It provides significant performance improvements over Enterprise COBOL V4. According to internal IBM test results, many well-structured, CPU-intensive batch applications have seen performance increases greater than 10 percent. Many numerically intensive programs have shown performance increases greater than 20 percent.
Users don’t have to optimize the entire application because optimized code works with non-optimized code. They can, therefore, achieve performance goals by selectively optimizing performance hotspots in an iterative and controlled way.

Maintaining Compatibility

Organizations can stage the upgrade effort to COBOL V5. They don’t need to recompile an entire application to use Enterprise COBOL V5. It’s source- and binary-compatible with older releases of COBOL compilers for System z. Most correct programs will compile and execute without changes, producing the same results. “Old” and “new” codes can interoperate within an application, communicating by static or dynamic calls.
Some old language syntax and options have been removed in the new compiler. They include supports for millennium language extension, label declaratives, non-reentrant programs above 16MB, OS/VS COBOL inter-operation, AMODE 24, and XMLPARSE(COMPAT).

Modernizing Applications

Application modernization promotes reuse of proven software assets. Enterprise COBOL V5.1 enables developers to methodologically modernize existing COBOL applications and integrate them with new ones running on modern infrastructures (e.g., Web, cloud or mobile). It allows organizations to deliver new enhancements quicker with lower risk and cost.
Enterprise COBOL V5.1 also supports Java 7 and provides users better control over generation of XML documents. It uses the z/OS XML parser, allowing parsing to be offloaded to specialty engines to reduce cost. It also provides support for unbounded tables and groups, which allows developers to enable top-down mapping of data structures between XML and COBOL applications.
As one IT architect with a North American health care company puts it, “XML is increasing in importance for our company, and IBM has enhanced XML support in Enterprise COBOL V5 with changes that allow us to take advantage of z/OS XML System Services to reduce the cost of COBOL-based XML processing.”


Improving Productivity

Enterprise COBOL V5.1 raised the total size of all data items in a working-storage or local-storage section to 2GB, and the maximum size of an individual data item to 999,999,999 bytes. New built-in functions are added to improve programmability of UTF-8 applications, thus improving COBOL applications’ ability to process UTF-8 data in DB2 and XML documents.
The new release continues to allow debugging of optimized production code. Debug information is stored in the NOLOAD segment of the program object. As a result, the footprint of the executable won’t increase. To improve application maintainability, a new floating comment indicator (*>) is added for developers to use anywhere in the program-text area to indicate that the text on the line is a comment line.
“COBOL V5 emphasizes IBM’s commitment to our existing COBOL application base while improving its interoperability with our open systems components, explains Michael A. Todd, a software architect for a multi-national financial services company. “IBM has addressed a number of architectural limits within the previous compiler, thereby extending the life of my extensive COBOL inventory. They continue to improve XML capabilities and UNICODE support, enabling us to reuse existing components.”
Additional benefits offered by Enterprise COBOL V5.1 include:
  • Reduced administration costs. The latest version supports a new level of z/OS System Management Facilities (SMF) tracking, allowing users who implement subcapacity tracking to reduce their administrative overhead.
  • Support for the latest middleware. It provides up-to-date programming interfaces for latest CICS, DB2 and IMS. The integrated DB2 and CICS coprocessors enable strong integration with the latest middleware technologies.
  • Support for modern development tools and solutions. By supporting the ecosystem of COBOL development tools from IBM and ISVs, it enables companies to expedite migration and application-performance tuning with IBM’s latest continuous integration solution.

COBOL Commitment

Enterprise COBOL for z/OS V5.1 affirms IBM’s commitment to COBOL customers on System z. With the new optimization framework, it delivers compelling performance improvement. Because it maintains source and binary compatibility with older COBOL releases, many applications may gain performance simply by recompiling the code.
Going forward, the new optimization framework will provide a solid foundation for delivering a release-to-release performance improvement roadmap for COBOL on System z. Therefore, it’s important to stay current with latest COBOL technology. Organizations can’t maximize their return on investment unless they stay current with latest compiler technology.

DUO claimt diverse schades na KPN-storing

Fouten in z/OS Catalogs uitwijkmainframe legden applicaties plat

05-03-2014 11:55 | Door Rik Sanders | Lees meer artikelen over: Disaster recovery, Mainframes, Replicatie, System z | Lees meer over de bedrijven: KPN, ministerie van Onderwijs, Cultuur en Wetenschap | Er zijn nog geen reacties op dit artikel | Permalink
De technische storing in het IBM-mainframe, die op zondag 27 oktober 2013 diverse systemen bij de Dienst Uitvoering Onderwijs (DUO) platlegde, heeft verschillende soorten schades veroorzaakt. Pas na afronding van de inventarisatie van de aard en de hoogte van de totale schade zal DUO bij ict-infrastructuurbeheerder KPN een mogelijke schadeclaim neerleggen. Ook vindt er nog een evaluatieonderzoek plaats naar de storing.
Dit blijkt uit de beschikking die het ministerie van Onderwijs, Cultuur en Wetenschap (OC&W) opstuurde, naar aanleiding van het door Computable opvragen van documenten over de storing op basis van de Wet Openbaarheid van Bestuur (WOB). Daarin wordt gesproken over een complicerende factor bij het bepalen van de aard en hoogte van de schade, namelijk dat er sprake is van 'verschillende soorten schades met elk hun eigen kenmerken'.

Extra handmatig werk

De storing in het mainframe legde vanaf eind oktober 2013 tot 4 november het studiefinancieringssysteem en het portaal Mijn DUO plat. Ook het personeelszakensysteem van Raet zou er uit hebben gelegen. Studenten konden door de storing niet inloggen op Mijn DUO. Medewerkers van de uitvoeringsorganisatie van OC&W waren niet in staat het studiefinancieringssysteem te raadplegen als studenten belden, mailden of langskwamen. Wel konden studenten via de website van DUO een aantal formulieren downloaden om bijvoorbeeld wijzigingen in de studiefinanciering door te geven.
De storing leidde tot veel extra werk in de uitvoering bij DUO, omdat medewerkers wijzigingen van klanten handmatig moesten noteren en, nadat de storing was opgelost, in het systeem moesten invoeren. DUO laat, sinds de systemen weer in de lucht zijn, het portaal Mijn DUO zwaar monitoren.

Catalog-fout

DUO event map Mainframe-verstoring
DUO event map Mainframe-verstoring
Volgens KPN is de storing veroorzaakt door een samenloop van twee omstandigheden: een fout in de replicatie van data tussen twee datacenters in combinatie met een geplande uitwijktest. De storing zorgde ervoor dat systemen niet benaderbaar waren en applicaties op die systemen niet gebruikt konden worden.

Uit een vrijgegeven WOB-document blijkt dat na het opstarten van het mainframe op de uitwijklocatie er een verstoring in het opslagsysteem plaatsvond. Daardoor ontstond er een inconsistentie in de z/OS Catalogs. Deze catalogs, die aangeven waar op welk volume een dataset is opgeslagen, verwezen in een aantal gevallen naar verkeerde locaties. Vervolgens werd het 'shared mainframe platform' onbruikbaar en konden de klantapplicaties niet meer beschikbaar worden gesteld (zie afbeelding).

Complex

DUO proces uitwijktest Mainframe
DUO proces uitwijktest Mainframe
Maar wat precies de oorzaak is geweest, wil DUO noch KPN aangeven. KPN, die als ict-infrastructuurbeheerder optreedt voor DUO, spreekt van een complexe storing. Een woordvoerder van het DUO stelt: 'De oorzaak was een fout bij KPN waarvan niet aannemelijk is dat die zich zal herhalen.'
Computable deed vervolgens een beroep op de WOB en vroeg de documenten op die te maken hebben met de storing. Het ministerie van OC&W, waar het DUO onder valt, wees het verzoek grotendeels af. Slechts drie van de 34 geïnventariseerde documenten zijn openbaar gemaakt, waaronder een persbericht en de pdf's 'DUO event map Mainframe verstoring' (zie afbeelding hierboven) en 'DUO proces uitwijktest Mainframe' (zie afbeelding hiernaast).

Afwijzen

Het ministerie staat op het standpunt dat openbaarmaking van de andere 31 documenten 'de positie van DUO bij de contractuele afhandeling van de storing en de bepaling van (de hoogte van) een mogelijke schadevergoeding zou kunnen verstoren.' De juridische dienst van OC&W voert ook nog aan dat openbaarmaking van de betreffende documenten DUO onevenredig zou benadelen, omdat de uitvoeringsorganisatie met KPN afspraken heeft gemaakt over de woordvoering omtrent de storing. 'Openbaarmaking zou ingaan tegen de afspraken rond de woordvoering en tevens de onderhandelingspositie van DUO kunnen verstoren en zo onevenredig kunnen benadelen', schrijft de dienstdoende jurist.
Wat OC&W met dit laatste punt bedoelt, wordt verder niet duidelijk gemaakt. Computable tekent bezwaar aan tegen deze beschikking.



Read more: http://www.computable.nl/artikel/nieuws/overheid/5020424/1277202/duo-claimt-diverse-schades-na-kpnstoring.html#ixzz2v5ZxVYE9

Wednesday, January 1, 2014

The Mainframe Comet, Burning Brightly: Reassessing the Big Iron Outlook

 

Posted by on in Mainframe   


If you’re reading this blog it means Earth wasn’t destroyed by the comet ISON. Phew! As we reflect on the fascinating path of that extraterrestrial wonder, it seems like a good time to introduce you to another “comet” we have in our sights when it comes to data centers and the mainframe.
Recently, we completed a guide on data center performance strategies and cost savings. In my discussions with recipients of the guide (which you can download for free here) and at presentations over the last year, I’ve made reference to the “mainframe comet.”
In essence, the view we take when forecasting the mainframe environment over the next 3-to-5 years resembles a comet. At the “tail end” of the comet, you’ve got a majority of customers who are looking to move some level of their data functions off the mainframe, equaling about half of the overall mainframe capacity. Burning bright at the other end is another 10% of customers, who make up one-quarter of all capacity. These are the global financial institutions, government agencies, and any business truly dealing with massive data volumes on a daily basis (in the ballpark of more than 20K MIPS).
In whole, the comet looks like this:

But as more strategies and technologies take advantage of existing mainframe environments – or at least optimize processing and I/O on z/OS – I’m curious if our “comet” may burn brighter and longer than anyone anticipates.
Hybrid environments and the data center are pulling mainframes into cutting-edge use. IBM, for one, has made significant investments and R&D dedications with zEDC, zBX and other System z hardware, earning effusive praise from writers in that community. The customers I’ve talked to surrounding the performance guide are all about doing more with – not without – their mainframe.
So what does the future of the mainframe look like to you? White-hot at 20K MIPS and beyond? A clunky way to keep around data for the auditor that you expect to flame out like the comet ISON? Or do you expect it to burn brightly in some other way?
When it comes to this continually emerging and even divisive segment of data performance, we always want to hear from you. We also welcome your feedback on our complimentary “big iron” performance guide as well as your comments below or via Twitter and LinkedIn. 



Comments