JPH06296179A - Diagnostic system for lan interconnection node - Google Patents

Abstract

PURPOSE:To execute a unit test and a network test of each interconnected node as required under the control of a monitor node in a LAN of bus configuration with respect to the diagnostic system of the LAN interconnection node in which the monitor node and plural general nodes are connected to the network of bus configuration. CONSTITUTION:A host device 1 of a monitor node 2 applies multiple address communication of a unit test command to a network 5 to disconnect logically the mode node 2 from the network 5 and its own node is tested by a unit test means 1c and connected again to the network 5 after the end of unit test. Upon the receipt of a test command from the network 5, the host device 3 of the general node 4 uses an identification means 3a to identify the test and when the host device 3 identifies the command to be a unit test command, the device 3 disconnects its ow node from the network 5 and uses a unit test means 3c to execute the test and after the end of test, its own node is connected to the network 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is an L node in which a monitoring node and a plurality of general nodes are connected to a network composed of a bus.
The present invention relates to a diagnostic method for an AN connection node.

In recent years, in the fields of information processing, communication, etc.,
A system in which a computer, a terminal, and a module are connected by a LAN to exchange data with each other is generally used.

In the case of a LAN constituted by a bus, it is necessary for the functions of the respective devices to function normally inside in order for the system to operate normally, and at the same time the communication function with other devices operates normally. Because there is a need for such LA
In N, a unit test of each node and a test in an operation form are performed.

[0004]

2. Description of the Related Art FIG. 10 is an explanatory view of a conventional example. Figure 10
LAN with multiple nodes N1 to N6 connected to the bus
The upper devices (indicated by P1 to P6) such as processors and terminal devices are connected to the nodes N1 to N6, respectively. Each host device transmits / receives data to / from each other through a network configured by a bus via each node. In the network shown in FIG. 10, the node N1 which is one of the plurality of nodes is the monitoring node,
The other nodes N2 to N6 are general nodes.

When testing the function of each node connected to the LAN in such a configuration, the test of the function as a single unit (the function of a single unit not including the function of communicating with other nodes) and the node connected to the network are tested. A network test (operation mode test) including communication between the two is performed.

Since a conventional unit test of a node is performed by disconnecting each node from the network, an engineer visits the location of each node and sets a switch SW provided between each node and the bus to OFF. Then, after physically disconnecting the node from the bus, an independent test is performed on each node (including the host device), and the switch SW is turned on after the test is completed.

Further, in the case of the network test (operation mode test), it is not possible to perform the test at the same time when the unit test is performed, and by issuing a command from the host device P1 of the monitoring node N1 to each general node, online This is performed by sending and receiving data having a predetermined content between the nodes designated by the state of (operation mode) and checking whether or not the node operates normally without error.

[0008]

In a conventional LAN in which a plurality of nodes are connected by a bus, in order to perform a unit test on all the nodes, each node is individually disconnected from the network, and therefore a large number of nodes are to be tested. There was a problem that it took time and effort because the person in charge had to go to the location of the node and execute it in order. Moreover, when performing a unit test, since it is performed separately from the network, there is a problem that a network test cannot be performed to check whether it is normal when connected to the network even if it is normal as a unit. Therefore, there is a problem that the network test must be performed separately from the unit test.

The present invention provides a LAN connection node diagnostic method capable of executing a unit test and a network test of each connected node in a LAN constituted by a bus under the control of a monitoring node when necessary. With the goal.

[0010]

FIG. 1 is a block diagram showing the principle of the present invention. In FIG. 1, 1 is a superordinate device (Ps) of the monitoring node, 2 is a supervising node (Ns), 2a is a gate provided between the supervising node 2 and the network, 3 is a superordinate device (Pn: n Is 1, ... N), 4
Is a general node (Nn), 4a is a gate provided between the general node and the network, 5 is a network composed of a bus, and a plurality (n) of general nodes 4 and their host devices 3 are connected to the bus. However, it is omitted in FIG.

In the host device 1 of the monitoring node, 1a is a unit test instruction means, 1b is a network disconnection means, 1
c is a unit test means, 1d is a network reconnection means, 1
Reference numeral e is a network test instruction means, 1f is a network test means, and 1g is a test result collecting means. Further, in the host device 3 of the general node, 3a is a test instruction type identifying means, 3b is a network disconnecting means, 3c is a unit testing means, 3d is a network connecting means, 3e is a network testing means, and 3f is a test result notifying means. .

According to the present invention, a single diagnosis is simultaneously executed for all nodes connected to the network, and after that, all nodes are transferred to the operation mode to test the normality of the communication function in the operation mode. is there.

[0013]

When the test is executed, the unit test instructing means 1a is driven in the host device (Ps) 1 of the monitoring node (Ns) 2 so that all the general nodes (Nn) from the monitoring node 2 via the network 5 are driven. A simultaneous broadcast instructing the unit test to 4 is sent. Each general node (Nn) 4
Then, the received broadcast is identified by the test instruction type identifying means 3a of the higher-level device (Pn) 3.

When it is determined that the unit test instruction is issued, the network disconnecting means 3b is driven in the host device 3 of the general node 4 to turn off the gate 4a between the general node 4 and the network 5. A signal off is generated to disconnect this general node 4 (including the host device 3). After the disconnection, the unit test means 3c is activated, and the unit function test of the node is executed according to the preset unit test data (displayed as Dt) prepared in advance. The result of the unit test in this general node 4 is obtained by
If the test result is good (normal), the network connection means 3d is driven, the control signal on is generated, and the gate 4a is switched on.

On the other hand, the host device of the monitoring node 2 sends the simultaneous broadcast of the unit test instruction to the network, and then the network disconnecting means 1b is driven to drive the gate 2
A control signal off for turning off a is generated, and the monitoring node 2
Disconnect (including the host device 1) from the network. After that, like the host device 3 of the general node 4 described above, in the host device 1 of the monitoring node 2, the unit test means 1c is driven to perform a unit test of the node according to the unit test control data (Dt), and the test is performed. After the end, network reconnecting means 1d
And the control signal on is supplied to the gate 2a to reconnect the monitoring node 2 and the network 5.

When the host device 1 of the monitoring node 2 is reconnected to the network 5, the test result collecting means 1g is driven to collect the unit test results from the general nodes 4 in order, and the result is obtained. Detect failure of a single function based on. It should be noted that this test result collecting means does not necessarily have to be performed at this time, and may be performed after the completion of all the tests.

Subsequently, in the host device 1 of the monitoring node 2,
When the communication with each general node 4 is established, the network test instructing means 1e is driven and the simultaneous broadcast is sent to all the general nodes 4 via the network 5. Upon receiving this, the host device 3 of each general node 4 identifies the type by the test instruction type identifying means 3a and determines the network test means 3
drive e. Network test control data (Dm) is prepared in the network test means 3e, and a network test including data transmission / reception with another node is executed according to the other network test control data (Dm).

At this time, the host device 1 of the monitoring node 2 also performs a network test by the network test means 1f according to the network test control data (Dm) provided therein. The network test control data (Dm) provided in each node has a plurality of patterns (D1, D2.
...) is prepared, and by selecting different patterns of Dm in each node, it is possible to perform a test simulating a state closer to the operation mode. When the network test is completed, the test result collecting means 1 of the host device 1
The g collects the network test result from the host device 3 of each general node 4 and detects the failure of the network function from the test result.

[0019]

Embodiment FIG. 2 shows the hardware configuration of the embodiment. In FIG. 2, reference numerals 10, 20, 30, 40, and 50 are devices corresponding to the respective units represented by 1 to 5 in FIG. 1, and 10 is a processor (P
s), 20 is a ring bus interface unit (RB
U is a monitoring node (Ns), 30 is a processor Pn (however, P in this example is a higher-level device of the general node 4).
1, P2) and 40 are general nodes Nn (N1, N2), which are ring bus interface units (RBU),
50 is a ring bus.

In each of the processors 10 and 30,
11 and 31 are central processing units (CC), 12 and 32 are main memory units (MM), and 13 and 33 are channel control units (CH).
C), 14, 34 are common bus interface controllers (CBC). Further, test control data groups Dx are stored in the main memory devices 12 and 32 in the processors 10 and 30, respectively, and the single function test control data Dt and the network test control data Dm (where m = 1, 2, 3
··)It is included.

In the configuration of the embodiment of FIG. 2, the central processing unit (CC) 11 of the processor Ps of the monitoring node 20 is
Execute task A for performing a test, and execute general node Nn
(40) Central processing unit CC of each processor P1, P2
The reference numeral 31 executes the task B according to an instruction from the processor Ps of the monitoring node Ns.

FIG. 3 is a process flow of task A executed by the processor of the monitoring node, FIG. 4 is a process flow of task B executed by the processor of each general node, and FIGS.
Indicates the state of each device of the network in the change of each process.

The processing flows of FIGS. 3 and 4 will be described with reference to FIGS. 2 and 5 to 8. First, in the processor Ps of the monitoring node Ns, an instruction to activate the task B is issued to the processors Pn (n = 1, 2 in this example) of all general nodes (S1 in FIG. 3). Next, a unit test instruction is sent to task B of all processors Pn (at step S2). This state is shown in FIG. 2, and the unit test instruction generated by the task A of the CC 11 of the processor Ps is sent to the general nodes N1 and N2 through the route shown by the dotted line in the ring bus 50 by simultaneous broadcast. next,
The processor Ps of the monitoring node Ns logically disconnects its own node Ns and carries out a unit test based on the unit function test control data Dt in the main memory MM (S3 in FIG. 3). FIG. 5 shows a state in which a single function test is executed in each node.

On the other hand, the processors P1 and P2 of the general nodes N1 and N2, as shown in FIG.
Task B is activated by the instruction generated in the processing of S1 and enters an instruction waiting state (S12 in FIG. 4). At this time, when the unit test instruction by the simultaneous broadcast based on S2 of the task A of the monitoring node Ns is received, the test instruction type is identified (S13 in FIG. 4), and when it is found that the unit test is performed,
A unit test instruction is generated, the own node Nn is disconnected, and a unit test is performed using the unit function test control data Dt (Fig. 4).
S14). This state almost coincides with the time when the unit test is performed in the monitoring node Ns, and is shown in FIG.

After the unit node functional test, the general node Nn logically connects its own node Nn to the ring bus 50 and incorporates it into the network (S15 in FIG. 4), and the unit test result is the task of the processor Ps of the monitoring node. Notify A (S16 of FIG. 4) and return to the instruction waiting state (S12 of FIG. 4).

In the processing flow of the processor Ps of the monitoring node shown in FIG. 3, when the single function test is completed, the own node Ns is incorporated into the network (S4 in FIG. 3). At this time, the processor Ps confirms the normal connection including whether the other general node Nn is normally connected to the network. Next, the unit test results of each node Nn are collected (S5 in FIG. 3). Figure 6 shows the status of this unit test result collection.
Shown in.

After collecting the test results, it is determined whether or not the unit test results for all the nodes have been collected (S6 in FIG. 3). If all the node results have been collected, based on the collected unit test results,
It is identified whether there is any abnormally terminated node (S7 in FIG. 3).
If there is an abnormally terminated node, the process of fault notification (display or printout on a console for a maintenance person (not shown)) is executed to terminate the task A.

When there is no abnormal termination node, the network test instruction is notified to the task B of all general nodes Pn (S8 in FIG. 3). FIG. 7 shows a state in which a network test instruction is transmitted, and this instruction is transmitted to each general node by a route indicated by a dotted line by simultaneous broadcast. Subsequently, the processor Ps selects one of the network test control data Dm (D1, D2 ...) Held in the internal main memory MM and carries out a network test according to the selected data (FIG. 3). S9). The state of the network test is shown in FIG. 8. In this example, the processor Ps selects D1 as the network test control data, the processor P1 selects D3, and the processor P2 selects D1 to perform the test.

On the other hand, in the processing of S8 in FIG.
The network test instruction transmitted by the processor Ps of the monitoring node Ns is received by the processor Pn of each general node. At this time, the processor Pn is in the instruction waiting state (S12 in FIG. 4) and identifies the test instruction type by reception (at step S13). If the result of this identification indicates that it is a network test instruction, one of the network test control data Dm (D1, D2 ...) Is selected and a network test is performed (S17 in FIG. 4).

This state is shown in FIG. 8 as described for the processor Ps. When this network test is completed, the network test result is notified to the processor Ps of the monitoring node Ns (S18 in FIG. 4).

Returning to FIG. 3, when the network test is completed, the processor Ps collects the network test result of each general node (S10 in FIG. 3) and identifies the test result to determine whether there is an abnormally terminated node. Yes (S1 in Figure 3
1). According to this determination, if there is an abnormal end node, a failure notification is given and end processing is performed, and if there is no abnormal end node, the processing of this task A ends.

Next, a network test executed by the processing flows shown in FIGS. 3 and 4 will be described. FIG. 9 shows an example of the network test control data Dm, D1, D
The network test control data indicated by 2 and D3 are data selected by the processors Ps, P1 and P2, respectively, and when performing a network test, the processors are synchronized with each other and sequentially based on their respective network test control data. The test is run.

For example, in TEST0, the data D1 of the processor Ps, the data D2 of the processor P1, and the control data D3 of the processor P2 each perform a reset test.
After that, each processor executes TEST1 according to the respective control data, the processor Ps executes the data transmission test, and transmits the predetermined data to the network. At this time, the processor P1 performs a loopback test (testing of the loopback function in the node in the duplicated network) with TEST2 of the control data D2, and the processor P2 performs a data reception test with TEST1 of the control data D3. The predetermined data transmitted and received is received. Thereafter, similarly, the respective tests TEST2 to TESTn of the control data D1, D2 and D3 are sequentially executed.

As described above, by changing the pattern of the network test control data for each node,
It is possible to perform a test that simulates a state close to the actual network operation mode.

[0035]

According to the present invention, a single function test (diagnosis) and a network test (diagnosis), which have been conventionally performed separately, can be collectively performed on all nodes, so that the test time can be shortened.

Further, in the conventional network test, the same test was carried out for all the nodes. However, according to the present invention, a test simulating a state close to the operational form is possible.
It is possible to improve the failure detection function and improve the reliability of network diagnosis for operation and maintenance.

[Brief description of drawings]

FIG. 1 is a principle configuration diagram of the present invention.

FIG. 2 is a hardware configuration diagram of the embodiment.

FIG. 3 is a task A executed by a processor of a monitoring node
Is a processing flow of.

FIG. 4 is a task B executed by a processor of a general node
Is a processing flow of.

FIG. 5 is a diagram showing a state in which a single function test is executed in each node.

FIG. 6 is a diagram showing a state of collecting unit test results.

FIG. 7 is a diagram showing a state in which a network test instruction is transmitted.

FIG. 8 is a diagram showing a state of a network test.

FIG. 9 is an example of network test control data.

FIG. 10 is an explanatory diagram of a conventional example.

[Explanation of symbols]

 1 Superordinate device (Ps) of monitoring node 1a Single unit test instructing means 1b Network disconnecting means 1c Single unit testing means 1d Network reconnecting means 1e Network test instructing means 1f Network testing means 1g Test result collecting means 2 Monitoring node (Ns) 2a Gate 3 Higher-level device (Pn) of general node 3a Test instruction type identification means 3b Network disconnection means 3c Single test means 3d Network connection means 3e Network test means 3f Test result notification means 4 General node (Nn) 4a Gate 5 Network

Claims (4)

[Claims] 1. In a diagnostic method of a LAN connection node in which a monitoring node and a plurality of general nodes are connected to a network composed of a bus, a host device of the monitoring node broadcasts a unit test instruction to the network and monitors it. Logically disconnect the node from the network, test the own node with the unit test means, reconnect to the network after the unit test is completed, and the host device of each general node identifies by the identifying means when receiving the test instruction from the network. When it is identified as a unit test instruction, the local node is disconnected from the network and the unit test means executes the test.
A diagnostic method for a LAN connection node, characterized in that the self node is connected to the network after the test is completed. 2. The host device of the monitoring node according to claim 1, which instructs the general node to issue a network test instruction after the test of the unit test is completed, and the host device of the ordinary node uses the identification means to perform the network test instruction. A diagnostic method for a LAN connection node, characterized in that, when there is identified, a function test in an operation mode via a network is performed by a network test means. 3. The host device of each of the monitoring node and the general node according to claim 2, comprising unit test control data and network test control data including a connection pattern for network test. ，
The LA is characterized in that the LA test is executed according to each unit test control data, and the network test means carries out a test according to each network test control data.
N connection node diagnostic method. 4. The host device of the monitoring node according to claim 1, wherein the upper device of the monitoring node collects the test result of the general node after the end of the unit test and the network test result of the general node after the end of the network test. A method for diagnosing a LAN connection node characterized by: