

Summary
Enterprise computing used to be
delivered using dumb terminals which, while not function rich, provided OnLine Transaction Processing (OLTP) levels of service (Remember them!). In usage terms, OLTP delivered sub
second response times thus ensuring that data could be entered and retrieved in
real time.
Since the advent of the Personal
Computer and laterally the Web browser, there is a constant improvement in the
client interface, however, this has resulted in severe degradation, in many
cases, of response times. The user experience is adversely impacted through
screens that take an age to appear. This is due primarily to two different
reasons. The first is that the amount of data to be delivered to the GUI for
display can take seconds or longer to be delivered over relatively slow links. The
second is that the functionality included in the GUI is so complex, existing
Web 2.0, Ajax or other frameworks can take seconds to complete the image.
The SOA
Gateway and CaptainCasa provide an Enterprise
Application Infrastructure using proven Service Oriented Architecture standards
to solve both of these issues and to enable organizations to return to sub
second response times without compromising on the functionality in the GUI.
About the Author
John
Power is a driving force behind the development of the SOA
Gateway and Managing Director of Risaris Limited. With over 25 years experience in the software
industry, John has delivered complex integration projects in Software AG, Delta
Airlines, Boston University and Morgan Stanley.
Contents
V.1.122/01/09
In the earlier years of online
computing, huge effort was put into reducing the amount of data being sent
backwards and forwards between display devices and their back end servers. This
was done as the speeds at which the devices were connected was extremely slow
and the bandwidth available on each of these connections had to be shared with
multiple devices. IBM recognised and addressed this many years ago with their
3270 protocol.
While not pretty by any means and
restrictive in means of availability of friendly user interface components the
3270 protocol was very efficient from communication perspective: the amount of
data for each client request/response was limited by the 80x24 character
screens that were sent from the server to the client. This was in particular
completely independent from the application processing on the server side –
whatever happened there, the communication to the frontend would never exceed
the 80x24 character data volume per roundtrip.
With the advent of the Internet, speeds
have increased exponentially so the same rules don’t apply....or do they?
Unfortunately the type and amount of data to be sent backwards and forwards to
display devices such as Browsers today has also increased exponentially. As
such, organizations today are sending very large amounts of data up and down
relatively larger pipes, however, the same issue still applies; if the increase
in data is commensurate with the increase in capacity, the same issue arises.
This is most noticeable in countries
where broadband speed and availability is relatively slow compared to the rest
of the world but also noticeable on the Internet. Consider that the speed at
which the data is delivered is subject to the slowest link on the Net-work
being used and we’re down to the lowest common denominator. So payload sizes do matter.
Apart from data volume perspective and
from client speed perspective, more and more people realize that the
HTML-technology (even though enhanced with JavaScript elements) does not meet
the user requirements for heavily used applications: the rendering speed of
HTML/JavaScript is slow compared to more native technologies, the availability
of power-user-features (hot keys, fully keyboard driven, …) is missing, the
possibility to integrate sub-devices (scanners,
…) is difficult. - HTML/JavaScript based user interfaces are very
dominant in the area of casual usage & anonymous user scenarios because of
their zero-installation approach. But they have problems meeting the
requirements of heavy-usage, operational scenarios.
CaptainCasa and the SOA Gateway provide
an ideal infrastructure to address these problems head on.
Providing an Enterprise Application
Infrastructure using this toolset provides many different benefits to the
business:
·
User who are heavily using applications daily, so called ‘power users’,
will be provided with a fully featured Graphical User Interface (GUI) for these
applications. This will provide all the benefits of simplicity that a good GUI
can bring.
·
Consistent response times of less than one second will enable these
users to be far more productive. In fact various studies and reports, many
going back to the early 80s, have shown that as response times increase, total
transaction volume decreases and errors increase.
·
Users satisfaction will improve as they can accurately predict how long
it will take to complete their tasks due to the consistency in the response
times.
·
The solution works in tandem with existing systems thus users on legacy
systems will see the same data as users on the new GUI system. This facilitates
a slow conversion for all users to the new system or both can continue to work
in tandem as the business requires.
·
Users, particularly in remote locations, will see a level of service
they can only have dreamed of.
·
Home workers working remotely can also simply use a standard Internet
connection to the office and see similar response levels to their office based
colleagues who may be running on a network many times faster than their home
network.
·
The tools provided avoids the bug budget integration spends that have
been a factor in such efforts in the past.
·
The solution can be implemented in small steps thus avoiding the big
bang approach and the requirement to commit large amounts of money and
resources up front. This means that benefits can be show in a relatively short
period of time.
A major factor prior to the start of any
project today is the Return on Investment (ROI) and how it can be measured. The
Enterprise Application Infrastructure described provides a number of tangible
points where the ROI can be measured:
·
Each business transaction will normally consist of one or more online
IT interactions. Thus the number of business transactions completed in a day
can be measured prior to the introduction and following the introduction of the
new infrastructure. As there is a value to each business transaction, it is
possible on a daily, weekly or monthly basis to determine what the value of the
change is to the business due to an increased number of transactions.
·
A well designed GUI interface can drastically cut the costs involved in
training new operators of the software, thus new employees can be up, running
and transacting business for an organization more quickly.
·
Multiple language support in the GUI ensures that the application is
written once and a language file translated to provide as many different
languages as are required by an organization. This leads to savings in
implementation in the first place but also in the fact that a single
application can service employees operating in multiple languages.
·
Many organizations in the past have upgraded network connections in
order to try to improve the response times to their users. Implementing this
Enterprise Application Infrastructure can potentially save such upgrades, which
can be expensive, or may even facilitate the removal of a previous upgrade to
save on costs.
There are also a number of intangible returns which, while not always
measurable directly, will certainly have an impact on the business.
·
Employees empowered to do their jobs well with these technologies will
generally be happier employees thus reducing the rates of staff turnover. This
leads to better trained staff, improved morale and so on.
·
Customers of the organization will be able to transact business with an
organization’s operators more quickly due to the quicker and more consistent
response times. This leads to happier customers and happier customers tend to
return to do more business.
While it is clear that all of the above
will be incredibly beneficial to the business, there are still many challenges:
·
In any existing organization, online applications will already exist to
run the business from day to day with existing GUI or non GUI style interfaces.
·
The likelihood is that these online applications will continue to
support the business for some time to come as it is rarely possible to go with
a ‘big bang’ approach. Therefore parallel running for some period of time will
be required.
·
The new GUI must be implemented seamlessly with the existing systems so
that any transaction executed using the new GUI will also be reflected to the
older online interface and vice versa.
The reuse of existing data and business
logic is the key to successfully implementing a new GUI for an application:
·
The new GUI implementation must interface with the existing data so
that the data upon which the business is built is available to any existing
systems as well.
·
It must reuse existing business logic as duplication of this logic in
different places is likely to lead to different results depending on which
online system is used.
·
The existing system and new online implementation must continue to run
in parallel until it is clear that both systems produce the same end result.
·
Clean interfaces between the new and existing systems will lead to
flexibility to change the interface as required and resilience when they are
changed.
·
Existing staff can generally move seamlessly to the new GUI solution
due to their familiarity with the older online interfaces that were used.
The first part of the
Enteprise Application Infrastructure solution, and the key to the massive
improvements that can be brought to the opeartion of the GUI is represented by
the following architecture diagram.

The user will start
their application in their favourite browser as before and the CaptainCasa run
time, running in a Tomcat or other compatible server, will deliver the initial
image. The key differentiator from this point in is that as changes are made in
the browser, only the items that have changes are transmitted via the GUI
interface to CaptainCasa. CaptainCasa then builds the output to the browser but
again only sends the changes to the browser which has the following impact:
·
The payload size for each
input and output operation is approximately 1-2K with this architecture
compared to “much more data” with a similar solution using HTML for example.
·
Intelligence in the browser
ensures that data is only transmitted when this makes sense. In other words,
data is not transmitted backwards and forwards each time the mouse is clicked
on the browser GUI image.
The second part of the Enterprise Application Infrastructure is
the novel way that applications and data are made available to the CaptainCasa
run time as per the following architecture diagram:

Using the SOA Gateway,
the organization’s existing business logic and data is made available to the
CaptainCasa run time using standards based Web Services This has the following
benefits:
·
Existing business logic in
COBOL, Natural, CICS and so on can be wrapped and made available as a Web
Service in minutes using the SOA Gateway tools.
·
Existing data in databases
like DB2, ADABAS, Oracle, Sybase etc. can be also wrapped and made available as
a Web Service in minutes using the SOA Gateway tools. This is a Create Read
Update Delete (CRUD) based service so data may be updated and accessed using
these services.
·
The CaptainCasa development
environment can import the Web Services Description Language (WSDL) from these
SOA Gateway services and create a default GUI for these new services.
·
The developer may then use
the CaptainCasa tooling to modify and beatify the screen layout, include
configuration to ensure valid data is entered at the GUI and to deploy to the
CaptainCasa run time server.
·
The GUI may then immediately
be used to start driving the application.
·
An additional key advantage
to this architecture is that the CaptainCasa run time will only drive the back
office Web Service when the input data has been validated against the rules
configured using the CaptainCasa development tools.
The architecture allows for a number of interesting deployment
options as follows.

In the above
configuration, the CaptainCasa server can run within a Cloud instance thus
taking advantage of the elasticity that Cloud computing provides for peak
loades on the server.

The existing data and business logic
could be migrated to a Cloud instance thus enabling the backend and front end
systems to take the benefit of Cloud computing.
·
The Enterprise Application Infrastructure described in this document
provides a capability to have rich GUI applications with the benefits of sub
second, OnLine Transaction Processing levels of service unavailable with other
infrastructures.
·
This can be done even though some users will have relatively slow links
to the server machines.
·
The tooling offered by the infrastructure can ensure that results and
benefits can be seen in weeks instead of months or years.
·
The infrastructure can work alongside existing deployments, be they GUI
or green screen, for as long as is necessary from a business perspective.
·
All of this can be achieved with minimal initial outlay based on the
usage based model adopted by the infrastructure.
The
SOA Gateway is a cost effective software tool to:
Access data faster...
It enables access to data from a wide range of
database languages (ADABAS, MySQL, DB2, VSAM, Oracle etc.) without server side
code, or expensive middleware.
Access
business logic easier...
The SOA Gateway enables easy access and re-use of
valuable business logic available in CICS, COBOL, C, NATURAL and many other
languages and environments.
The
SOA Gateway is developed by integration
specialists Risaris Limited.
For a free trail of the SOA Gateway, please visit: http://www.soagateway.com/html/registration_form.php
This document is distributed for
information purposes only and does not form part of or constitute an
agreement with Risaris Ltd. Although Risaris Ltd. uses reasonable efforts
to include accurate and up-to-date information in this document, Risaris
makes no warranties or representations as to its accuracy. Risaris Ltd. may
also make improvements and/or changes to this document at any time without
notice. The various approaches outlined in this document are put forward in
good faith, but it remains possible that individual results may vary. For
that reason and in accordance with standard practice, readers are
encouraged to test any materials developed on the basis of this paper
before putting them into productive use.
More
whitepapers at:
http://www.soagateway.com/html/industry_papers.html
See
Terms & conditions at www.soagateway.com
Ó
Risaris Limited 2009.
