JPH0456445A - Test system for network system - Google Patents

Abstract

PURPOSE:To improve the efficiency of system test by using one test program for an object all layers constituting a network system so as to conduct a test between optional layers with a unified method. CONSTITUTION:Two computers 1 are connected by a local area network 3 via a transceiver cable 2 and the system is made up of five protocol layers 4-8. One test program 25 exists on a highest layer 4, which accesses directly each of inter-layer communication interfaces (IF) 26-30. Thus, protocol tests 14-18 between layers of a same rank, IF tests 19-24 between adjacent layers and also a protocol test and IF test of a single layer and between optional layers are conducted. Thus, a trouble of employing a different test program from layers or a special equipment is not caused and the test is conducted unifiedly.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a testing method for a network system having a layered structure.

(Prior art) For example, in Fig. 5, two computers (1) are connected via a transceiver cable (2) via a LAN (3), and five layer protocols (4), (5) are connected.
.

It has a network system consisting of (6), (7), and (8). In addition, test programs for each layer (9) (1o), fll), (12), (
13) exists as a higher-level application program for each layer and has an interface for each layer. With this test program, protocol tests between peer layers (14), (15), (18),
(17), (18) and interface tests between adjacent layers (19), (20),' (21)
, (22).

(23) and (24) can be performed.

For example, in a protocol test between layers (6), a test program (12) is used to actually perform communication to confirm that communication can be performed according to the protocol of layer (6).

In addition, the interface test between layer (5) and layer (6) is performed by actually using the interface between layer (5) and layer (6) in the test program (11) to confirm that the test program operates normally. Good rubbing. Therefore, testing can be performed on layer (6) as long as one knows the protocol and interface of layer (6) and how to use a test program. In other words, the tester must check the parts other than the layer to be tested (other layers, LAN, upper application).
It was difficult to understand the network as a whole system, whether it was possible to test the target layer without special knowledge of the network.

(Problem to be solved by the invention) As described above, the conventional test method uses a different test program for each layer, so when a failure occurs in the network system, the failure occurs in the same layer. One had to learn how to use different test programs for all layers in order to determine whether the

In addition, since the system is constructed by stacking multiple protocols and interfaces, it is difficult to understand the operation of the entire system.In particular, the higher the layer, the harder it becomes to see the lower layer protocols and interfaces. It is almost impossible to perform primary analysis of the failure, that is, to isolate the failure.

This invention was made to solve the above problems, and allows testing of all layers using one upper-level application program, and also allows system testers to understand the operation of the entire network system. The purpose of this study is to obtain a network system testing method that facilitates primary analysis of failures.

(Means for Solving the Problems) A network system testing method according to the present invention includes a test program that is located at the highest layer among a plurality of protocol layers constituting a network system and can directly access communication interfaces between each layer. Create an information table regarding all the target layers and all communication interfaces, and start the test by referring to the above information table with the commands manually input from the predetermined data manual means.
means for specifying a layer for outputting termination and intermediate progress, and for analyzing input commands and setting parameters for a communication interface to start test execution;
It has a means for collecting data that has passed through an intermediate progress output layer or an end layer and status information of the network system at that time, and editing and outputting it to a predetermined output means.

(Function) In the present invention, by using one test program that targets all layers constituting a network system, tests between arbitrary layers can be performed in a uniform manner, making it more efficient for system testers. Provide a good testing environment.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. 1st
The figure is a schematic diagram of an embodiment of the network system testing method according to the present invention, and is drawn in correspondence with FIG. 5. Two computers (1) are connected via a LAN (3) via a transceiver cable (2), and five protocol layers (4), (5), (5).

This is a network system consisting of (7) and (8).

One test program (25) on top layer (4)
There is direct access to the communication interface between each layer (26), (27), (28),
(29), (301).This allows the conventional protocol test between peer layers (14), (
15), (16), (17), (18) and interface tests between adjacent layers (19),
(20).

(21), (22), (23), (24)
Of course, it is also possible to perform protocol tests and interface tests on individual layers or between arbitrary layers.

Referring to FIG. 2, the implementation method of this embodiment will be explained in more detail. FIG. 2 is an enlarged view of one computer (1) in FIG. 1. In the same figure, (211), (27)
, (211), (29), and (30) are the communication interfaces (31) and (32) of each layer from the test program (25).

(33), (34), and (35) can be directly accessed, but it is assumed that this interface also includes a test control interface for each layer.

The tester (36) inputs commands to the manual command analysis/execution program (4o) via, for example, a file (38) in the fixed disk device (37) or a keyboard (39). Although the format of the Macondo to be input is not particularly specified in this embodiment, the human information included in the command includes the layer at which the test is started, the interface and its parameters at which execution starts in the start layer, the layer at which the test ends, etc. (Process Sl in Figure 4). The human command analysis/execution program (40) analyzes input commands using the human command information table (41). This table is, for example, a table as shown in Figure 3, which includes the layer (protocol) name (42), the number (43) that identifies the layer (protocol), and the information that the layer (protocol) provides to the upper layer (protocol). The number of communication interfaces (including test control interfaces) (44), each interface salt (45) and the format of the human parameter required by that interface (46)
, an execution start interface flag (47) to know the interface to start the test, an end layer flag (48) to know the layer to end the test, an address where log trace information of each layer is stored (49), etc. It is included.

It is assumed that these pieces of information have been set in advance in the table (41), and that the execution start interface flag (4
7) and the end layer flag (48) columns are all set to OFF as initial values.

The human command analysis/execution program is based on this table (
Referring to Figure 3), the manual command is checked and an error is returned if it is incorrect. When the manual input is correct, turn on the execution start interface flag of the interface that starts the test, and turn the end layer flag of the test end layer to OFF.
N, and further convert the manual command into the manual parameter format (46) of the execution start interface. Then, it notifies the start layer and end layer through the test control interface (50), (51), sets parameters in the execution start interface and starts execution (52).
) (Step S2 in FIG. 4(a)).

Each layer (4), (5), (6), (7)
, (8) is a communication interface for each layer (including,
Ink face for test control) (26), (27)
, (28), (29), (30), and catch/response routines (53), (54), (55) that catch the arrival of data and notify the test program (25) of the results as necessary. )
, (56) (57), and a routine that collects log traces, status information, etc. of each layer and outputs them to the specified user space (address) (58), (5
9).

(60), (61), and (62) (in this embodiment, log tracing, status information collection method, output format, etc. are not specified).

For example, the layer (8) that is notified of the termination layer (51) via the test control interface (30)/
Response routine (57) Holds information that the own layer is the end layer. When the human data (63) passes through the communication interface (35) of the layer (8), the catch/response routine (57) catches the arrival of the data and informs the test control interface that the data has arrived at the end layer. At the same time, the content of the arrived data is returned via the communication interface (30) (step 53 in FIG. 4(a)).

Then, the command analysis/execution program (40) takes in the returned data and outputs the results to an output device such as a printer (66), display (67), or file (38) (see Fig. 4(a)). Process S4).

On the other hand, the test program (25) includes a log/trace/status information collection function for each layer (58) (5!])
A maintenance tool program (65) that edits and outputs the information collected by , (60), (61), and (62).
). The information collected for each layer is recorded in a predetermined area, and the address where the information is recorded is in the log/trace information recording address column (49) of the table (Figure 3). shows the address of that area. By referring to this address (49), the maintenance tool program (65) takes in any information of any layer and sends it to the link (66), the display (67), and the file (38) on the disk (37).
) etc. so that the examiner (36) can easily understand the results (Fig. 4(b)).

〔Effect of the invention〕

As described above, in this invention, all layers are tested using one test program, so there is no need to learn how to use a different test program for each layer or use special equipment. Tests can be conducted uniformly. In addition, since it is executed manually using simple commands and the results are edited and output, anyone who knows the interface between each layer can easily perform the test and analyze failures. Therefore, the number of people and time required for testing can be reduced, resulting in more efficient testing.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of a network system test method according to an embodiment of the present invention, FIG. 2 is a detailed diagram of this embodiment, and FIG.
The figure shows the contents of the table (41) in this embodiment, FIGS. 4(a) and 4(b) are flowcharts explaining the operation of this embodiment, and FIG. 5 is a schematic diagram of a conventional network system test method. It is. (1) is the computer, (4) to (8) are each layer, (14)
~(18) is a protocol test for each layer, (19) ~
(24) is an interface test between adjacent layers, (
25) is the test program of this embodiment method, (26) to (
30) is the interface between this embodiment method test program and each layer, (31) to (35) are the communication interfaces of each layer, (36) is the tester, (37) is the fixed disk device, and (38) is the file. , (39) is a keyboard device, (40) is a human command analysis/execution program,
(41) is a table. In each figure, the same reference numerals indicate the same or equivalent parts.

Claims (1)

[Claims] An information table regarding all layers and communication interfaces to be tested is created in the test program, which is located at the highest layer among the multiple protocol layers that make up the network system and can directly access the communication interfaces between each layer. The command input from the input means is referred to the above information table to specify the layer for starting, ending, and outputting the progress of the test, and also to analyze the input command and set parameters for the communication interface that should start the test execution. and a means for collecting data passing through an intermediate output layer or an end layer and status information of the network system at that time, and editing and outputting the collected data to a predetermined output means. Network system test method.