Data processing: structural design – modeling – simulation – and em – Emulation
Reexamination Certificate
2000-03-02
2004-12-14
Thomson, W. D. (Department: 2123)
Data processing: structural design, modeling, simulation, and em
Emulation
C703S013000, C702S182000, C702S186000, C709S223000
Reexamination Certificate
active
06832184
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention relates generally to a simulation method and system for testing computer networks, and more particularly to a method and system for generating local area network (“LAN”) traffic for multiple simulated client workstations.
BACKGROUND OF THE INVENTION
Computers and computer networks currently provide important advantages to enterprises and individuals in today's society. Moreover, with the advent and ensuing popularity of the Internet and the World Wide Web (“Web”), there is a tremendous increase in volume and usage of networked computer systems. Consequently, the explosive growth of computer networks has necessitated a need for larger servers to handle the network traffic. Presently, different implementation paradigms are available for handling this increase in volume. For example, some developers in the computer industry are implementing larger single servers to handle the network traffic while others are including a greater number of duplicated, relatively small servers to handle the increase in the network traffic volume.
Typically, as part of a development cycle, computer systems and software applications are tested by utilizing a simulation test tool, designed to emulate a real run-time environment in order to test whether the specific computer system or application meets the various design criteria. In both of the above-described server implementations for meeting the high demand of network e) traffic, a complete and thorough test requires that the servicing computer, e.g., a server comprising a single machine or multiple machines, be tested at full load. Failure to execute a complete test inevitably results in a failure in the system when a full load is eventually impressed upon the server during the run-time. Accordingly, to effectively test large and complex distributed applications and/or server applications, simulation of a realistically large client base is needed. Currently existing simulation tools provide simulation at application level. However, simulation at the application level yields an application specific simulator. Simulation at the port level is an improvement but still fails to exercise many client-specific paths in the system under test. Thus, a general purpose simulator that seeks to provide a high fidelity simulation, for example, a simulation at level
2
, the data link layer of the protocol stack, is highly desirable. The data link layer is the lowest protocol stack level where each individual client has a unique client address.
Other simulators are currently available for testing computer networks to withstand a large volume of traffic.
FIG. 1
is a diagram
100
illustrating a typical prior art simulation setup. The serving system, i.e., a system under test
102
is driven by a set of simulation driving systems, S
104
. These driving systems
104
are controlled by a central machine, simulation controller
106
. All of these machines share a common connection medium
108
such as a local area network or LAN. A plurality of such configurations may be present in order to achieve the total bandwidth necessary to drive the serving system. There may be multiple controlling systems or the controlling system may have connections to multiple LANs.
Typically, the existing prior art simulators fall into two broad categories: 1) keystroke stuffers; and 2) protocol exercisers. A “keystroke stuffer” utilizes the actual application on the driving client. The driving vehicle is a program which retrieves stored scripts of keystrokes, mouse movements, and/or other user inputs to provide input to the actual client application to drive the application. The application then directs the traffic through a client protocol stack to the system under test
102
.
FIG. 2
is a diagram of the seven layer Open Systems Interconnection (“OSI”) protocol stack model. Each of the layers represents a function that must be performed to effect communication between different machines. The lowest layer in the model is the physical layer
202
. The functions within the physical layer
202
include setting up, maintaining, and deactivating physical circuits U) between systems. The most notable physical layer interfaces include IEEE 802.2 and IEEE 802.3. The next layer, i.e., layer
2
, in the OSI model is the data link layer
204
, which is responsible for transferring data over the physical circuits or the channel between systems. The functions of the data link layer
204
include dividing data into frames in order to transfer the frames to another system across the physical medium. The data link layer provides for the synchronization of data to delimit the flow of bits from the physical layer. The data link is a point-to-point link between two devices that are directly connected together.
The next layer, layer
3
, is the network layer
206
. The network layer
206
provides inter-network services such as the network routing and the communications between networks. The network layer
206
handles multiple point-to-point links in the case where frames are transmitted across multiple links to reach their destination. Internet Protocol (“IP”) in the Transmission Control Protocol/Internet Protocol (“TCP/IP”) suite is a network layer protocol.
The next layer, layer
4
, is the transport layer
208
. The transport layer
208
provides end-to-end accountability of data transmitted as streams of packets. The transport layer functions include monitoring data flow to ensure proper delivery of data between source and destination. It provides for error correction and for data fragmentation and reassembly. TCP is a protocol layer protocol.
The next layer up, layer
5
, is the session layer
210
. The session layer
210
provides for control and synchronization in exchanging data between users. For example, dialogues may be used for check and recovery of data transfer. The next layer up, layer
6
, is the presentation layer
212
. The presentation layer
212
functions include formatting data for display or presentation. In this layer, codes and encryption in data are interpreted and formatted for presentation. The next layer up, layer
7
, is the application layer
214
. This layer is responsible for supporting end-user applications such as file transfer, electronic message exchanges and terminal session.
Importantly, it should be understood that all layers of the above-described OSI model are not necessarily present in all protocol stacks. That is, the OSI protocol stack is a reference model to provide standardized logical decomposition of network into layers for communications between systems. For example TCP/IP does not have a complete level
4
in the protocol stack because IP does not guarantee delivery of data nor will it detect missing packets, although there is a reassembly function in TCP/IP that serves a portion of the function typically found at level
4
. Instead, for TCP/IP, reliable delivery is left up to the driving application at level
7
. Thus, a keystroke stuffer is, effectively, simulation above level
7
, the application layer of the protocol stack.
The second type of prior art simulators is a “protocol exerciser,” which provides a somewhat more efficient method for producing load. This method involves simulation at level
7
of the protocol stack. This type of simulator will have knowledge of a particular application, for example, the File Transfer Protocol (“FTP”). This simulator will open ports directly with the system under test rather than having the client application do so and drive traffic to the system under test by conforming to the higher level protocol that is being tested, e.g., FTP, Hypertext Transfer Protocol (“HTTP”), Open Database Connectivity (“ODBC”), Simple Mail Transfer Protocol (“SMTP”), etc. The protocol exerciser provides an efficient way to generate load because all extraneous functions of the application such as redrawing screens, etc., do not need to be performed at the client. A significant drawback of this type of load generation, however, is that this simulator is protocol specific. If
Bleier, Jr. William H.
Blue Dale E.
Brady Kevin J.
Jones Jeff R.
Maurer Max M.
Kinnaman, Jr. William A.
Scully Scott Murphy & Presser
Thomson W. D.
LandOfFree
Intelligent work station simulation—generalized LAN... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Intelligent work station simulation—generalized LAN..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Intelligent work station simulation—generalized LAN... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3310101