One sales person at IBM is no doubt having a very good week. The company just announced the inking of a modernization deal with Travelport, a company that processes travel transactions for more than 60,000 travel agencies worldwide.
Through this multi-million-dollar agreement, Travelport hopes to double its transactional processing capabilities. The deal includes upgrading to the IBM z/Transactional Processing Facility (zTPF), as well as WebSphere, Rational, Tivoli and a boatload of IBM hardware.
The two companies already had a working relationship before this storage and middleware expansion. Last year, Travelport finished upgrading the underlying infrastructure of its global distribution system with IBM. Through this project, the company consolidated all of its data center operations into Atlanta, Georgia. Travelport must have been pleased with the results.
This is a pretty big deal for IBM. Travelport handles travel transactions that involve any of 420 airlines, 88,000 hotel properties, 25 rental car companies and a host of other entities across the travel industry. Representatives say Travelport’s system processes up to 1.6 billion messages a day for transactions like airline and hotel bookings.
There are a variety of caching techniques to be considered in grid, cloud and other types of distributed computing architecture for analytics. Among these, the object data base can show some advantages said Carl W. Olofson, Research Vice President, IDC. In fact, some of the early object data base companies are positioning their wares for cloud computing.
“It is definitely an alternative in some data caching uses, particularly those where you need to share the data across a wide variety of systems, and especially where there are long running transactions involved,” said Olofson.
With the ODB technique, a Java developer, for example, can embed the interaction of the code and the data and use the natural functions in the Java language to create the application, Olofson noted.
Some viewers may be skeptical of an ODB resurgence – in the 1990s, object data bases were often touted as an alternative to relational bases, and they never found nearly as much use. But today’s ODB survivors are more likely to be muted in their pronouncements. ODBs for new distributed apps are likely to be called supplements to RDBs and data caches in special cases.
A lot of data architectures – both familiar and new – are vying today for consideration by development teams pursuing new apps that provide very fast analytics on very large sets of data. Object databases (ODBs) and persistent object stores represent one of these data architectures. Continued »
Noting here that Ed Roberts, inventor of Altair personal computer, died at 68. There is no one person that can claim the invention of the personal computer. But the Altair inspired many others. Roberts is far less known than Bill Gates and Steve Jobs – both paid tribute at the time of his passing.
Inventor of Altair personal computer, dies at 68 – BBC
When exposing Web services to entities external to your organization, it is easy enough to roll your own security and management policies – if you don’t have a lot of services. As B2B integration points increase in number, however, many find it useful to handle the various areas of policy management from a single utility. Most companies would probably prefer their developers spend more time on business logic than infrastructural concerns.
The folks over at SOA Software recently released Policy Manager 6 with revamped security federation and policy management interoperability. The company has also refactored the product to run on top of the highly-extensible OSGi framework. Continued »
Retrieving data demands both time and compute power of an application, and that demand can grow when data is housed in disparate data centers, as in a cloud or grid computing environment. Distributed data caches can make the data a little easier to find. A distributed data cache can act as an intermediary storage layer, holding frequently used data so that the application doesn’t have to constantly query multiple databases. Continued »
The pendulum swing toward lighter Web frameworks – sometimes called ‘implementations’ – probably will not swing in one direction infinitely. Within the plethora of features that made the original Web application servers ‘fat’ are elements needed by some applications. Fat has its place. Continued »
IBM recently opened up its public IBM Cloud for test and development after a beta period. The offering relies on Red Hat’s KVM-based hypervisor for its underlying virtualization system.
This new “Smart Business Development and Test on the IBM Cloud,” is not to be confused with CloudBurst, Big Blue’s private cloud product that lives behind the corporate firewall. The company says its new public cloud (or “commercial cloud” as IBM terms it) can, however, work in conjunction with CloudBurst.
It is not terribly surprising to see a large vendor using open source technologies as the underpinning for a cloud infrastructure. Amazon uses open source Xen for virtualization and Google – well its love of open source goes without saying. What may surprise some is that, well, it’s IBM. Continued »
Some have projected Java to be a likely language of the cloud. The thing is, within the Java community, there seems to be a vocal group that sees the Java Virtual Machine – rather than the Java language – as the main contributor to future cloud architecture. Continued »
Business Process Management (BPM) implementers need to consider the varied ways existing enterprise service buses (ESBs) use to communicate, according to Gartner analyst Roy Schulte. Continued »