Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1997-11-21
2001-02-06
Beausoliel, Jr., Robert W. (Department: 2785)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C709S219000, C717S152000
Reexamination Certificate
active
06185701
ABSTRACT:
TECHNICAL FIELD
This invention relates to Internet testing tools and, more particularly, to systems and methods for client-based web server application verification and testing.
BACKGROUND OF THE INVENTION
In the early stages of the phenomenal growth of the Internet, fairly simplistic web sites predominated, permitting the web “surfer” to simply and quickly access information displayed on a graphical user interface screen, and then log off. Such early web sites were relatively unsophisticated. However, as the richness and utility of the Internet began to cause its well-chronicled and explosive growth, the potential of the Internet began to be recognized by web site designers, and commercial enterprises. Thus, not only did web sites increase in complexity in terms of such things as hot links and nesting of pages, but additional richer features began to appear, such as the ability to post and get data downloaded to files and the dynamic changing of the web site as a function of user interaction or profiles whereby custom web pages would appear.
As the complexity and need for dependability of applications running on web servers increased (such as with the advent of commercial on-line banking and shopping transactions), the need arose to be able to rigorously test these more complex applications for their integrity and robustness. Turning to
FIG. 1
, depicted therein is a simplified view of the Internet environment in which such an application resides, including the mechanisms typically utilized prior to the subject invention for effecting such testing and verification of the Internet application. Typically a web server
10
is provided on which one or more web applications such as the banking application
12
are running. The web server
10
is interconnected to end users
16
and a core controller
14
, the latter, in turn facilitating access to information, for example, of a financial institution
18
reposing on its respective databases
20
. In such a typical Internet application as the illustrated banking application, this system thereby enables huge numbers of end users to quickly, efficiently, and in a secured fashion, transact their financial business with the financial institution over the Internet.
With the advent of these more serious and complex applications of the Internet and with the corresponding increasingly dire consequences which might arise from misbehavior of the various components of
FIG. 1 and
, in particular, those consequences arising from faults with the application
12
and server
10
, a need arose to rigorously test these web server applications
12
. One can readily appreciate the disaster awaiting a commercial concern such as an airline ticket reservation company or bank handling large volumes of transactions and dollar amounts on a daily business, should it be found that various logical program design flaws exist in the application
12
. It is precisely for these reasons, that a great deal of research and technology development was directed at ways to test these applications for integrity prior to their introduction as a live application into the Internet. The prior manner of doing so is further illustrated in FIG.
1
. In effecting such testing, it was conventional to provide for one or more testers
22
who would manually exercise the given application
12
, often employing “screen scrapers” so as to capture and store in storage
24
, various screen images delivered to the tester
22
from the application
12
as the tester(s) traversed the various links of the application.
Several problems are associated with this approach to testing of web server applications. First, the process was extremely slow due to the associated screen draws and saving of GUI data as well as the manual nature of the process. This manual aspect of the testing further led to failure to fully test the application due to links missed either by mistake or the sheer number of such links involved, failure to exhaust all possibilities of data input/output, inability to retest and compare data, and the like.
Attempts to automate the screen scraping tasks yielded some improvement but did not address even more serious flaws with the approach to validating Internet applications. The data which was being tested and saved often was browser dependent and really resulted in merely testing the browser front end and the images being returned thereby. However, what was really needed was an efficient and thorough way to exercise all of the APIs and code itself of the application running on the web server. In a typical application such as a banking application there may be literally hundreds of APIs and reference pages associated therewith for performing various functions such as login, account summaries, and so forth. This “back-end” of a given web server application therefore has associated with it an immense amount of program logic and data “gets” and “puts” such that it would be extremely desirable to be able to test and verify not only whether images from a particular browser are as expected, but that this data being entered and returned as a result of traversing the web site and its associated links was in fact correct. Thus a system was needed which, in an automated fashion, could test and verify the logic and data and associated myriad permutations and combinations of APIs and reference pages associated with a web server application rather than merely testing and debugging the browser and associated GUIs per se. Such a system was needed which, unlike the prior systems, could avoid saving data such as these GUI images which were not critical to testing this “back-end” logic and information developed by the server application.
As the sophistication of web server applications increased, yet an additional problem surface in facilitating the testing of the web applications. The ability to fashion dynamically configuring web server application pages soon developed in the maturation of the web which could build HTML pages on the fly. While it was known in the art to fashion static test cases to test web pages either in a manual or automated fashion, a difficulty arose in essentially providing for dynamic test cases which could test these changing web pages. Even in a static sense, making changes from a prior test (such as adding fields, changing fonts, moving images, resizing screens, etc.) would break test pages, and necessitate the rewriting of such tests. Moreover, when these changes could occur in a dynamic way as is currently the state of the art, this even compounded further the ability to test and verify web server applications in a meaningful way.
A related problem to this dynamic nature of web pages is that it was highly desirable to perform repeated tests on server applications whereby comparisons could be made to prior data. However, in the case of prior art, the data which was being saved (e.g., screens) was not the data critical to testing of the integrity of the underlying data. Rather, it was associated with the browser per se and thus was not the data and logic of paramount importance in verifying the web application itself. Thus, it was highly desirable to provide a mechanism for verifying data from repeat visits by comparison to previously acquired data.
In summary then, a need existed to be able to efficiently and automatically request, capture, store, and verify data returned from web servers disassociated from the particular browser itself, wherein such data related more importantly to the actual underlying data and logic of the server application. An ability was needed to run in an automated fashion to avoid errors and lack of thoroughness associated with manual traversal of web sites. Further, an ability was needed to discard data now critical to such verification and testing. Moreover, an option was highly desired for saving and reusing this returned data for the testing of subsequent transactions. Still further, it was highly desirable to avoid use of a conventional browser per se, so as to avoid the browser interfering with the tests which were of real interest
Marullo Frank Robert
Randall, Jr. Donald Husted
Beausoliel, Jr. Robert W.
Carwell Robert M.
Elisca Pierre E.
International Business Machines - Corporation
LandOfFree
Automated client-based web application URL link extraction... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automated client-based web application URL link extraction..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automated client-based web application URL link extraction... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2593392