IBM brightened up its Pulse conference in Las Vegas last week with a number of interesting announcements. Interesting for two reasons: it gives users some new useful products and it hints at the direction that IBM’s thinking is taking it.
In classic business-speak we found out that by providing clients with greater visibility into the performance of their businesses and control over the management of business processes, IBM helps them improve their ability to respond to changes, and gain greater command over their hardware, software, people, and information assets. Which, to be honest, actually sounds useful for many organizations. If the banks had only known that they were loaning so much money to so many people who couldn’t ever pay it back, we wouldn’t be in our current financial crisis!!
Here’s a brief list of the announcements.
The Tivoli Service Automation Manager software automates the design, deployment, and management of services such as middleware, applications, hardware, and networks.
The Tivoli Key Lifecycle Manager software helps organizations simplify the life-cycle of encryption keys by enabling them to centralize, automate, and strengthen security through key management processes.
The ProtecTIER Deduplication Appliance helps eliminates redundant copies of the same data, reducing it to a single instance of the data, and eliminating duplicative copies. Thank you Diligent Technologies.
There are enhancements to the IBM XIV Storage System.
There is Full-Disk Encryption on IBM System Storage DS8000s, which can help increase security. By combining the DS8000 with its full-disk encrypting drives and the Tivoli Key Lifecycle Manager, the DS8000 security solution can secure data and securely erase data from disk drives that are being retired or used for a different purpose.
IBM Internet Security System’s new data security services can help prevent information loss through: network extrusion prevention; implementing and managing encryption solutions to protect data even if lost or stolen; securing data more effectively; monitoring and controlling the use of external storage devices for storing and transporting data; and enhancing the security of inbound and outbound e-mails.
The InfoSphere Warehouse for System z beta software allows customers to more easily and cost effectively design and run a data warehouse that supports business intelligence applications.
The IBM Tivoli Monitoring for Energy Management software brings automation to the management and reporting of energy consumption by non-IT assets, eg an office building air conditioning system.
Quite a lot that will be useful to the business community. And IBM is clearly looking for a single point for the monitoring and control of much more than just the mainframe.
Monday, 16 February 2009
Monday, 9 February 2009
Mainframe on a stick!
In the past, I’ve looked at mainframes that aren’t really mainframes! You know things like PSI’s Open Mainframe, Fundamental Software’s FLEX-ES, Hercules, and a couple of other options. I’ve even mentioned the “L” word – litigation. What I haven’t mentioned is the IBM System z Personal Development Tool (zPDT).
Anyway, for those of you who’ve tuned in late, here’s the story so far…
PSI’s Open Mainframe was able to run z/OS, but IBM wouldn’t licence z/OS for the PSI machines. Then IBM sued PSI, and PSI counter-sued IBM. For Fundamental’s customers, the wrangling with PSI meant that IBM didn’t re-licence Fundamental’s FLEX-ES. Similarly with Hercules, IBM wouldn’t licence products for their machines. At the end of last June, IBM bought PSI.
Now, I’ve been arguing for ages in this blog that IBM should support mainframe software running on platforms other than mainframes because it would be a way of seducing a whole new group of potential customers to the IBM way of doing things. A middle-sized user has a room full of servers doing a mixture of work. At the moment, the best choice for them is to get a couple of big servers and run VMware’s Virtual Machine software, and get rid of most of their other boxes. But what if they could have a go using z/OS instead. They may very well find it not only does all the work for them, but it also introduces them to a culture of back-ups and archives and logging, that’s an auditor’s dream. And that’s where the Open Mainframe, FLES-ES, and Hercules options came in.
I suggested previously that IBM might be planning to destroy the competition and then bring to market its own version of the product. And in a stealth-like way, that’s perhaps what it’s done. Way back in April 2007, it released internally the IBM System z Personal Development Tool (zPDT).
What do we know about it? It seems this z/Architecture emulation software runs on a ThinkPad/T60 (an Intel-powered PC) running SUSE, and on AIX (on POWER chips). It requires a USB hardware-key plugged in to enable the CPU. This is an IBM 1090 (or the IBM System z Personal Development Tool Adapter – zPDTA).
So what’s IBM’s plan? I suggest that the internal release gave them the opportunity to beta test their software and ensure it was pretty much bug free. But what does IBM do next? Does it e-mail users of Open Mainframe, FLEX-ES, and any other emulation package in town, and tell them about a new opportunity? Does it phone up all the other software vendors and ask them whether they’d like to test their mainframe software on a PC? Or does it start touring the PC conferences and exhibitions trying to sell this new product?
I’m not sure, but I can remember many years ago, coming across just such an IBMer (Mark Cathcart, in fact) at a mainly PC-based show demoing the P/390 machine. And he seemed to be busy with potential customers all day. I would guess that there could be a lot of interest in a PC-based mainframe emulator.
I expect to hear much more about it during the coming year.
BTW, I followed Dougie Lawson’s advice on the IMS-L listserv and joined Twitter – I’m at http://twitter.com/t_eddolls. Now I just need someone to explain what I can do with it that I can’t do with Facebook, LinkedIn, Plaxo, etc, etc. And they could also explain why every time I do try to do anything I’m told that Twitter is over capacity!
And finally, don’t forget to download your free copy of the Arcati Mainframe Yearbook 2009 from www.arcati.com/newyearbook09.
Anyway, for those of you who’ve tuned in late, here’s the story so far…
PSI’s Open Mainframe was able to run z/OS, but IBM wouldn’t licence z/OS for the PSI machines. Then IBM sued PSI, and PSI counter-sued IBM. For Fundamental’s customers, the wrangling with PSI meant that IBM didn’t re-licence Fundamental’s FLEX-ES. Similarly with Hercules, IBM wouldn’t licence products for their machines. At the end of last June, IBM bought PSI.
Now, I’ve been arguing for ages in this blog that IBM should support mainframe software running on platforms other than mainframes because it would be a way of seducing a whole new group of potential customers to the IBM way of doing things. A middle-sized user has a room full of servers doing a mixture of work. At the moment, the best choice for them is to get a couple of big servers and run VMware’s Virtual Machine software, and get rid of most of their other boxes. But what if they could have a go using z/OS instead. They may very well find it not only does all the work for them, but it also introduces them to a culture of back-ups and archives and logging, that’s an auditor’s dream. And that’s where the Open Mainframe, FLES-ES, and Hercules options came in.
I suggested previously that IBM might be planning to destroy the competition and then bring to market its own version of the product. And in a stealth-like way, that’s perhaps what it’s done. Way back in April 2007, it released internally the IBM System z Personal Development Tool (zPDT).
What do we know about it? It seems this z/Architecture emulation software runs on a ThinkPad/T60 (an Intel-powered PC) running SUSE, and on AIX (on POWER chips). It requires a USB hardware-key plugged in to enable the CPU. This is an IBM 1090 (or the IBM System z Personal Development Tool Adapter – zPDTA).
So what’s IBM’s plan? I suggest that the internal release gave them the opportunity to beta test their software and ensure it was pretty much bug free. But what does IBM do next? Does it e-mail users of Open Mainframe, FLEX-ES, and any other emulation package in town, and tell them about a new opportunity? Does it phone up all the other software vendors and ask them whether they’d like to test their mainframe software on a PC? Or does it start touring the PC conferences and exhibitions trying to sell this new product?
I’m not sure, but I can remember many years ago, coming across just such an IBMer (Mark Cathcart, in fact) at a mainly PC-based show demoing the P/390 machine. And he seemed to be busy with potential customers all day. I would guess that there could be a lot of interest in a PC-based mainframe emulator.
I expect to hear much more about it during the coming year.
BTW, I followed Dougie Lawson’s advice on the IMS-L listserv and joined Twitter – I’m at http://twitter.com/t_eddolls. Now I just need someone to explain what I can do with it that I can’t do with Facebook, LinkedIn, Plaxo, etc, etc. And they could also explain why every time I do try to do anything I’m told that Twitter is over capacity!
And finally, don’t forget to download your free copy of the Arcati Mainframe Yearbook 2009 from www.arcati.com/newyearbook09.
Monday, 2 February 2009
Arcati Mainframe Yearbook 2009 user survey
The Arcati Mainframe Yearbook 2009 has been available for download free from www.arcati.com/newyearbook09 for just over a week. During that time we have had lots of e-mails saying nice things about the Yearbook including: "Excellent work", "Arcati's Mainframe Yearbook is probably the best publication about IBM Mainframe at the moment", and "It is very useful". It is full of interesting stuff, but particularly interesting are the results of the user survey.
Respondents all completed a survey on the Arcati site between the 3 November 2008 and the 5 December 2008. 38% were from Europe and 43% from North America, with 16% from the rest of the world and 3% declining to respond.
50% of the respondents worked in companies with upwards of 10,000 employees worldwide, while 15% had 1001 to 5000 staff. 12% of respondents had 0-200 staff and 9% had 5001-10000. In terms of MIPS 46% of respondents (down from last year's 41%) had fewer than 1000 MIPS installed, 23% fell into the mid-sized category between 1000 and 10,000 MIPS, and 31% were at the high end.
Looking at MIPS growth produced some interesting results. Larger, more mature businesses (above 10,000 MIPS) were mostly experiencing some growth, but none were growing by more than 25%. Sites in the 500-10,000 MIPS range were typically experiencing growth in the 10-25% range. Sites below 500 MIPS were generally experiencing growth of less than 10%, but only two sites reported zero growth or negative growth. The mainframe market does appear to be quite fragmented with competitive pressures at the lower end of the mainframe market.
With the environment and environmental issues getting so much coverage in the media these days, the survey asked whether IBM's recent green initiatives on things like power consumption and cooling had made the mainframe more or less attractive. Nearly two-thirds (62%) said that IBM's green initiatives made no difference at all. No-one said that the initiative made the mainframe less attractive. 29% felt it made the mainframe a little more attractive, and 9% felt it made the mainframe a lot more attractive.
There are a huge number of other gems of information in the survey. A wake-up call in its conclusion states that it is essential that businesses understand where the [mainframe] system's real strengths lie, and how it can be integrated most effectively with other IT resources.
How true!
Respondents all completed a survey on the Arcati site between the 3 November 2008 and the 5 December 2008. 38% were from Europe and 43% from North America, with 16% from the rest of the world and 3% declining to respond.
50% of the respondents worked in companies with upwards of 10,000 employees worldwide, while 15% had 1001 to 5000 staff. 12% of respondents had 0-200 staff and 9% had 5001-10000. In terms of MIPS 46% of respondents (down from last year's 41%) had fewer than 1000 MIPS installed, 23% fell into the mid-sized category between 1000 and 10,000 MIPS, and 31% were at the high end.
Looking at MIPS growth produced some interesting results. Larger, more mature businesses (above 10,000 MIPS) were mostly experiencing some growth, but none were growing by more than 25%. Sites in the 500-10,000 MIPS range were typically experiencing growth in the 10-25% range. Sites below 500 MIPS were generally experiencing growth of less than 10%, but only two sites reported zero growth or negative growth. The mainframe market does appear to be quite fragmented with competitive pressures at the lower end of the mainframe market.
With the environment and environmental issues getting so much coverage in the media these days, the survey asked whether IBM's recent green initiatives on things like power consumption and cooling had made the mainframe more or less attractive. Nearly two-thirds (62%) said that IBM's green initiatives made no difference at all. No-one said that the initiative made the mainframe less attractive. 29% felt it made the mainframe a little more attractive, and 9% felt it made the mainframe a lot more attractive.
There are a huge number of other gems of information in the survey. A wake-up call in its conclusion states that it is essential that businesses understand where the [mainframe] system's real strengths lie, and how it can be integrated most effectively with other IT resources.
How true!
Monday, 26 January 2009
The Arcati Mainframe Yearbook 2009
The 2009 edition of the celebrated Arcati Mainframe Yearbook is now available to view or download (free of charge) from www.arcati.com/newyearbook09.
It includes a mainframe strategy section with a selection of articles by industry gurus and vendors on topics such as: IBM z/OS VWLC software pricing optimization; The credit crunch: the impact on System z (aka mainframes); and Staying on top of Enterprise Extender.
For many people the highlight each year is the mainframe user survey. This illustrates just what's been happening at users’ sites. It’s a good way for mainframers to compare what they are planning to do with what other sites have done. I will be looking at some of the survey highlights next week.
The full vendor directory section gives contact details and other useful information in a summary format for all major vendors, consultants, and service providers in the z/OS environment.
There’s also a media guide for IBM mainframers. This lists information resources, publications, and user groups for the z/OS environment.
Next comes the glossary of terminology, which contains a large number of definitions of mainly mainframe-related terms.
The final section contains technical information – z10 model tables; mainframe hardware timeline 1952-2009; and mainframe operating system development.
In order for the Yearbook to be free, it is supported by sponsors and advertisers. This year's sponsors were Software Diversified Services (SDS), Value-4IT Ltd, and zCOST Management.
To see this year's Arcati Mainframe Yearbook, click on http://www.arcati.com/newyearbook09.
Now that the Yearbook is complete... if anyone needs a good technical writer etc, please contact me at trevor@itech-ed.com.
It includes a mainframe strategy section with a selection of articles by industry gurus and vendors on topics such as: IBM z/OS VWLC software pricing optimization; The credit crunch: the impact on System z (aka mainframes); and Staying on top of Enterprise Extender.
For many people the highlight each year is the mainframe user survey. This illustrates just what's been happening at users’ sites. It’s a good way for mainframers to compare what they are planning to do with what other sites have done. I will be looking at some of the survey highlights next week.
The full vendor directory section gives contact details and other useful information in a summary format for all major vendors, consultants, and service providers in the z/OS environment.
There’s also a media guide for IBM mainframers. This lists information resources, publications, and user groups for the z/OS environment.
Next comes the glossary of terminology, which contains a large number of definitions of mainly mainframe-related terms.
The final section contains technical information – z10 model tables; mainframe hardware timeline 1952-2009; and mainframe operating system development.
In order for the Yearbook to be free, it is supported by sponsors and advertisers. This year's sponsors were Software Diversified Services (SDS), Value-4IT Ltd, and zCOST Management.
To see this year's Arcati Mainframe Yearbook, click on http://www.arcati.com/newyearbook09.
Now that the Yearbook is complete... if anyone needs a good technical writer etc, please contact me at trevor@itech-ed.com.
Monday, 19 January 2009
The quick and the dead
I thought this week I would talk a little bit about Agile software development. If you’ve not come across it before, and I’m sure you probably have, it refers to a number of similar methodologies for software development, for example Scrum. Why is it worth talking about? Well the answer is to do with the way anything developed can be rapidly updated in a controlled way making the best use of teamwork.
Basically, Agile software development is a way of managing a project that involves numerous small improvements to a piece of software rather than starting with a massive long-term plan and working towards it. It’s sort of like the ant approach to solving a problem. Lots of little actions take place by a number of different individuals and at the end of the project some massive piece of software has been created (or in the ants’ case, a massive ant hill!).
Each of these small iterations is worked on by a team, and they work their way through the standard software development life-cycle. The time for these iterations is usually between a week and a month. So each increment goes through planning, requirement analysis, design, coding, unit testing, and acceptance testing phases.
The technique is quite popular with programmers because there is usually little emphasis on written documentation, although there is face-to-face communication, and the team, which is usually between five and ten people, typically contains a customer representative. In fact that person is usually the one creating any written documentation.
At the end of each iteration, customers and stakeholders will review the progress so far and can then re-evaluate priorities. So this is why the technique is described as agile. The original plan is not written in stone and new choices can be made at any time during the project to ensure that customers maximize their return on investment.
Team members usually get to know each other quite well and know what each one is working on. Regular meetings of the whole team can help to identify any problems and the team as a whole can work on solving them. Meetings are held with everyone standing up – which focuses the mind and stops meetings dragging on for long periods of time.
Agile software development originated in the 1990s. The Agile Alliance (www.agilealliance.org) was formed by a number of organizations in the early 2000s to promote Agile software development. Agile methodologies like to see themselves as being adaptive, and they contrast that to other methodologies that are predictive. Typically, engineering methods tend to try to plan out a large part of the software process in great detail over a long span of time. The problem with this approach is that things may change between the plan being originally devised and the software being finally delivered. This means that they try to resist change. The advantage that agile methodologies offer is that they embrace change. They can adapt to changes and include them in the final software.
A second difference is that Agile methodologies are people-oriented rather than process-oriented. This means that the final process will be one that is usable by people rather needing people to learn to use the process.
So Agile software development is designed for scope creep – mission creep – that process whereby something was created for one purpose and then (while it’s there!) has other things required of it.
So why call the article the ‘quick and the dead’? I was metaphorically suggesting that Agile computing is quick (and very much alive) and older monolithic strategies might be dead in the water. What do you think?
Basically, Agile software development is a way of managing a project that involves numerous small improvements to a piece of software rather than starting with a massive long-term plan and working towards it. It’s sort of like the ant approach to solving a problem. Lots of little actions take place by a number of different individuals and at the end of the project some massive piece of software has been created (or in the ants’ case, a massive ant hill!).
Each of these small iterations is worked on by a team, and they work their way through the standard software development life-cycle. The time for these iterations is usually between a week and a month. So each increment goes through planning, requirement analysis, design, coding, unit testing, and acceptance testing phases.
The technique is quite popular with programmers because there is usually little emphasis on written documentation, although there is face-to-face communication, and the team, which is usually between five and ten people, typically contains a customer representative. In fact that person is usually the one creating any written documentation.
At the end of each iteration, customers and stakeholders will review the progress so far and can then re-evaluate priorities. So this is why the technique is described as agile. The original plan is not written in stone and new choices can be made at any time during the project to ensure that customers maximize their return on investment.
Team members usually get to know each other quite well and know what each one is working on. Regular meetings of the whole team can help to identify any problems and the team as a whole can work on solving them. Meetings are held with everyone standing up – which focuses the mind and stops meetings dragging on for long periods of time.
Agile software development originated in the 1990s. The Agile Alliance (www.agilealliance.org) was formed by a number of organizations in the early 2000s to promote Agile software development. Agile methodologies like to see themselves as being adaptive, and they contrast that to other methodologies that are predictive. Typically, engineering methods tend to try to plan out a large part of the software process in great detail over a long span of time. The problem with this approach is that things may change between the plan being originally devised and the software being finally delivered. This means that they try to resist change. The advantage that agile methodologies offer is that they embrace change. They can adapt to changes and include them in the final software.
A second difference is that Agile methodologies are people-oriented rather than process-oriented. This means that the final process will be one that is usable by people rather needing people to learn to use the process.
So Agile software development is designed for scope creep – mission creep – that process whereby something was created for one purpose and then (while it’s there!) has other things required of it.
So why call the article the ‘quick and the dead’? I was metaphorically suggesting that Agile computing is quick (and very much alive) and older monolithic strategies might be dead in the water. What do you think?
Monday, 12 January 2009
CICS Explorer
IBM introduced a SupportPac containing the new CICS Explorer Eclipse-based graphical tooling interface for CICS on the 5 November 2008. But what is it?
Basically, it’s a new Eclipse-based systems management environment for application programmers, system programmers, and other operational staff. The idea behind it, I guess from IBM’s point of view, is to provide a modern interface to CICS, which, as you know, as been around since 1969, and is currently at Version 3.2 (although second time around because it reached Version 6 then started again!).
CICS Explorer is highly-customizable, and can be used to integrate information from CICS and third-party tools, which means that users need have only one Master Terminal for all their software. IBM has made using it easier for beginners because it comes with integrated help, tutorials, and a direct connection to the CICS Information Center.
The IBM site at http://www-01.ibm.com/software/htp/cics/explorer/features/ lists its features including:
Why are modern and easy-to-use such important features? Well it goes back to my blog the other week about the ageing mainframe work force. This particular tool helps to fix the ageing-workforce problem in two ways.
Firstly, you can replace experienced and skilled mature staff with less-experienced and skilled staff, because it’s easy for them to access help from CICS Explorer. So there’s no need to replace retiring experts with younger experts, you just need to replace them with standard IT people.
The second thing is that staff who are familiar with Java can make use of Explorer’s Eclipse interface, and before they know it, they are experts at CICS! And Eclipse-based plug-ins can be integrated in CICS. The old and the new combined, making working with CICS sound youthful and dynamic and appealing to the younger Java programmers.
And because it can run in a browser, it can be run from anywhere. I’m currently picturing a die-hard CICS programmer from 20 years ago, fixing a problem using the browser on his phone, while attending a family party. Work-life balance and all that!
Definitely a product we’re going to hear more of in 2009.
Basically, it’s a new Eclipse-based systems management environment for application programmers, system programmers, and other operational staff. The idea behind it, I guess from IBM’s point of view, is to provide a modern interface to CICS, which, as you know, as been around since 1969, and is currently at Version 3.2 (although second time around because it reached Version 6 then started again!).
CICS Explorer is highly-customizable, and can be used to integrate information from CICS and third-party tools, which means that users need have only one Master Terminal for all their software. IBM has made using it easier for beginners because it comes with integrated help, tutorials, and a direct connection to the CICS Information Center.
The IBM site at http://www-01.ibm.com/software/htp/cics/explorer/features/ lists its features including:
- Manage secure connections to CICS runtime and CICS Tools from a single panel.
- Wide range of standard queries plus rich query editor so you can easily modify these or create your own.
- Much easier to follow a sequence of resource relationships.
- Data filtering helps to reduce the volume of data displayed.
- CICS Explorer can run in a browser.
- Etc.
Why are modern and easy-to-use such important features? Well it goes back to my blog the other week about the ageing mainframe work force. This particular tool helps to fix the ageing-workforce problem in two ways.
Firstly, you can replace experienced and skilled mature staff with less-experienced and skilled staff, because it’s easy for them to access help from CICS Explorer. So there’s no need to replace retiring experts with younger experts, you just need to replace them with standard IT people.
The second thing is that staff who are familiar with Java can make use of Explorer’s Eclipse interface, and before they know it, they are experts at CICS! And Eclipse-based plug-ins can be integrated in CICS. The old and the new combined, making working with CICS sound youthful and dynamic and appealing to the younger Java programmers.
And because it can run in a browser, it can be run from anywhere. I’m currently picturing a die-hard CICS programmer from 20 years ago, fixing a problem using the browser on his phone, while attending a family party. Work-life balance and all that!
Definitely a product we’re going to hear more of in 2009.
Monday, 5 January 2009
Enterprise WOA
I talked about mash-ups and the enterprise a couple of weeks ago. This time I’d like to take a look at WOA and how that fits into the enterprise.
So, like a good philosophy essay, let’s start with some definitions. What is WOA, or perhaps more simply, what do the letters of the acronym stand for. WOA stands for Web-Oriented Architecture. It’s a way of building large-scale Web services. It’s like SOA on steroids! Gartner have a sort of definition. They think of it as SOA plus WWW plus REST. And just to unpick that definition a little, it’s Service-Oriented Architecture (the one that we’ve all been hearing about for the past five years) plus the World-Wide Web plus REpresentational State Transfer.
SOA, as you’re probably well aware, is a software design that breaks down a problem into modular, distributable, shareable, and loosely-coupled components. The World-Wide Web part refers to Web-related technologies such as HTTP(S) (HyperText Transfer Protocol Secure), (X)HTML (eXtensible Hypertext Mark-up Language), RSS (Really Simple Syndication), Atom Publishing Protocol, RDF (Resource Description Framework), and others.
With REST, there are no objects or methods, rather, everything is a resource. In a RESTful system, an agent communicates with a resource using its location, a verb describing the operation, and a representation of the desired state.
SOA usually comes associated with SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language). Now the problem many people experience is that the WSDL specification can be difficult to understand, domain specific, and difficult to extend or adapt. In addition, SOAP messages can be complex and difficult to understand without some knowledge of the services involved.
The advantage of WOA is that it uses existing technologies. A WOA interface can be used from any Web-enabled programming language. WOA can be easily extended.
But, I hear you say, can I use this with my mainframe? The answer is yes. And the reason is that WOA is ready for this kind of usage. It is simple, scalable, and extensible. It makes use of all the benefits of HTTP – such as using PUT, GET, POST, and DELETE requests. And it promotes the use of semantically rich representations.
You may remember me saying that mainframe mash-ups can use REST. The use of REST is probably the big difference at the moment between traditional SOA and WOA – although, every diagram shows WOA as being a subset of SOA. And it’s not all good news about REST, there are certain areas where it’s not great – such as two-phase commit, messaging, and asynchronicity.
But having said that, the use of REST and SOA puts the focus on Web resources rather than services and it just might be the way to get SOA projects to a satisfactory conclusion.
So, like a good philosophy essay, let’s start with some definitions. What is WOA, or perhaps more simply, what do the letters of the acronym stand for. WOA stands for Web-Oriented Architecture. It’s a way of building large-scale Web services. It’s like SOA on steroids! Gartner have a sort of definition. They think of it as SOA plus WWW plus REST. And just to unpick that definition a little, it’s Service-Oriented Architecture (the one that we’ve all been hearing about for the past five years) plus the World-Wide Web plus REpresentational State Transfer.
SOA, as you’re probably well aware, is a software design that breaks down a problem into modular, distributable, shareable, and loosely-coupled components. The World-Wide Web part refers to Web-related technologies such as HTTP(S) (HyperText Transfer Protocol Secure), (X)HTML (eXtensible Hypertext Mark-up Language), RSS (Really Simple Syndication), Atom Publishing Protocol, RDF (Resource Description Framework), and others.
With REST, there are no objects or methods, rather, everything is a resource. In a RESTful system, an agent communicates with a resource using its location, a verb describing the operation, and a representation of the desired state.
SOA usually comes associated with SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language). Now the problem many people experience is that the WSDL specification can be difficult to understand, domain specific, and difficult to extend or adapt. In addition, SOAP messages can be complex and difficult to understand without some knowledge of the services involved.
The advantage of WOA is that it uses existing technologies. A WOA interface can be used from any Web-enabled programming language. WOA can be easily extended.
But, I hear you say, can I use this with my mainframe? The answer is yes. And the reason is that WOA is ready for this kind of usage. It is simple, scalable, and extensible. It makes use of all the benefits of HTTP – such as using PUT, GET, POST, and DELETE requests. And it promotes the use of semantically rich representations.
You may remember me saying that mainframe mash-ups can use REST. The use of REST is probably the big difference at the moment between traditional SOA and WOA – although, every diagram shows WOA as being a subset of SOA. And it’s not all good news about REST, there are certain areas where it’s not great – such as two-phase commit, messaging, and asynchronicity.
But having said that, the use of REST and SOA puts the focus on Web resources rather than services and it just might be the way to get SOA projects to a satisfactory conclusion.
Subscribe to:
Posts (Atom)