JPH09305438A - Fault information transfer system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To respond to a command even during the transfer of a report and to prevent useless transfer when the reception side of the report disables reception by transferring the response to the command from the transmission side of fault information and transferring the report of remaining fault information later. SOLUTION: A slave CPU transfers a report #1 of fault information to a main CPU and the main CPU transfers the received report #1 to an NMS. Continuously, the slave CPU transfers a report #2 to the main CPU. At such a time, the main CPU receives a command from the NMS before the transfer of the report #2 to the NMS. Then, the slave CPU receives the transfer of the command from the main CPU and preferentially transfers the response to the received command. Besides, when the response from the slave CPU is received, the main CPU preferentially transfers the response to the NMS. Thus, when the reception side of reports #1 and #2 disables the reception, the useless transfer can be prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fault information transfer system, and more particularly, it can respond to a command even while a report is being transferred, and can avoid unnecessary transfer when the report receiving side cannot receive the command. A fault information transfer method that can be performed.

[0002] An integrated services integrated digital network (ISDN) has been introduced as an infrastructure of an advanced information society, and various information networks have been constructed. As the scale increases and the functions of the devices that make it up become more sophisticated and complex, it is important to manage the network so that network resources can be efficiently operated and high quality of service can be maintained.

International Standards Organization (ISO)
Organization for Standardization) is promoting the standardization of network management architecture, but specifies the following network management functions.

(1) Fault Management (2) Configuration Management (3) Performance Management (4) Confidentiality Management (5) Accounting Management Of these, the present invention relates to fault management. The functions related to fault management include grasping and reporting of fault conditions, diagnosis support for display of fault locations, reporting of impact areas, activation of avoidance procedures, restoration confirmation test support, grasp of countermeasure status, and statistical processing of reported fault information. The present invention relates to grasping, reporting and displaying a fault situation.

As described above, a plurality of central processing units (CPUs) are provided because the functions of the network and the communication equipments forming the network are sophisticated and the structure is complicated.
Distributed monitoring is performed by and the monitoring result is transferred to the monitoring center. The equivalent of this monitoring center is a network management system (NMS: Net
work Management System). Therefore, a plurality of CPUs performing distributed monitoring need to transfer fault information to the NMS.

FIG. 10 shows an example of the configuration of a system to which the invention is applied. In FIG. 10, 1 is NMS, 2a and 2
b is a main shelf provided with a monitoring function unit of the communication devices forming the network, and 3a and 3b are communication function units of the communication devices forming the network.
It is a slave shelf that does not have a monitoring function unit. And
The main shelf is a main CP that mainly communicates with the NMS.
U21, slave CPU22 that monitors the slave shelf, master C
It is configured by a communication path 23 between the PU and the sub CPU. Further, 4 is a communication path between the communication device and the NMS, which is physically configured by a transmission path that constitutes a network. Further, 5 is a communication path between the slave CPU and the slave shelf, which physically extends the bus of the slave CPU.

If the main shelf is provided with a communication function unit, the monitoring is performed in the same manner as the slave shelf.
It should be done by the PU. FIG. 11 shows the configuration of the CPU.

In FIG. 11, reference numeral 211 denotes a processor, 2
Reference numeral 12 is a random access memory (RAM), 213 is a read-only memory (ROM), 214 is a display device installed as necessary, 215 is a bus, and 216 is a communication port. In addition, in FIG. 11, all the constituent elements are shown as one, but a plurality of necessary constituent elements may be provided.

Now, when the configuration of FIG. 11 is a main CPU,
Two communication ports are provided, one of which is connected to a communication path with the NMS (to be precise, connected to the NMS via a multiplexer and a transmission line of a main signal transmitted through the network) and the other. Is connected to the communication path with the slave CPU.

On the other hand, when the configuration of FIG. 11 is a slave CPU, a display device and a slave shelf to be monitored are connected to the bus 215, and one communication port is connected to a communication path with the main CPU. .

Therefore, in the configuration of FIG. 10, the slave CPU monitors the communication function unit (hereinafter referred to as a device) provided in the slave shelf. When the slave CPU detects a change point in the state of the device, the detection result is displayed on the display device connected to the slave CPU, transferred to the main CPU as a report, and further transferred from the main CPU to the NMS.

FIG. 12 is a diagram for explaining collection of device states and change point detection. The state table collects and stores the device states. This state table includes a previous value table that stores the previously collected states and a current value table that stores the newly collected information this time.

In FIG. 12, if the numbers in parentheses are the numbers of the monitoring points of the device, 1 is the failure occurrence, and 0 is the failure recovery, the previously collected states are the monitoring points (1) and (n). Is a failure, and the status collected this time is the monitoring point (1)
And (2) are obstacles. If the contents of the previous value table and the contents of the current value table are exclusive-OR'ed for each monitoring point, the change between the previous state and the current state can be known.

What is obtained from this is the change point detection table, and it can be seen that a change in state has occurred at the monitoring points (2) and (n). Since the number of this monitoring point is the same as the address of each table, if a value table is searched this time using the changed addresses (2) and (n) as a key, a failure occurs at monitoring point (2). However, it can be seen that the failure at the monitoring point (n) has been recovered.

The contents of this change point detection table are transferred to the NMS as a report. On the other hand, when the maintenance person on the NMS side wants to confirm the current state of the slave shelf, a command can be issued and a response can be requested. The response of the slave CPU to this command is the response. For example, NM
When the S side issues a command to confirm the current status, the contents of the current value table are transferred as a response.

Here, for reference, the structure of the command and the response or the report is shown. FIG. 13 is a command configuration example. In FIG. 13, the horizontal direction is the bit width and the vertical direction is the word width. Usually, the bit width is 8 bits, that is, 1 byte, so the word width is the number of bytes. This also applies to FIG.

As shown in FIG. 13, the command is composed of a header including an address for specifying the CPU, an instruction code, addresses of shelves and units, data, and a CRC code. The above data is significant when the command is intended for control, and is null data when confirming the current state.

FIG. 14 is a structural example of a response or a report. As shown in FIG. 14, the response or report is a header including an address that identifies the CPU, a code corresponding to the instruction code, the address of the shelf or unit, the presence or absence of a continuous response or report, and the length indicating the number of continuous responses or reports. , Data, CRC
It is composed of a code. Incidentally, whether it is a response or a report is displayed by a code corresponding to the instruction code.

As described above, the device status is monitored and the result is reported and the response is made in response to the request of the host device. However, the transfer of the change point information does not delay the monitoring of the device status. It is required to realize a fault information transfer method that can avoid the transfer of necessary change point information.

[0020]

2. Description of the Related Art FIG. 8 shows an operation of a conventional fault information transfer system which corresponds to detection of a single change point of a slave CPU.

The transfer procedure of the conventional fault information transfer method will be described below with reference to the reference numerals of FIG. S51. Monitor the device status and detect changes. (Here, it is assumed that the change point is detected.) S52. Create a report. The number of reports is N (N is a positive integer).

S53. Set the counter to 1. S54. It is determined whether the count value of the counter is larger than the number N of reports. S55. If the count value of the counter is not greater than N (No), the report is transferred.

S56. The count value of the counter is incremented by 1, and the process returns to step S54. Then, S54, S55, and S56 are repeatedly executed until the count value of the counter becomes larger than N.

During this period, if the command from the NMS is received, the received command is stored. S57. When it is determined in step S54 that the count value of the counter is larger than N (Yes), it is determined whether or not a command is received.

If no command has been received (No), a series of operations is ended. S58. If the command is received (Yes),
The response is transferred in response to the command, and the series of operations ends.

Since FIG. 8 shows the operation corresponding to the detection of one change point in a closed manner, a report is transferred, and a series of operations is terminated if the command is responded to. However, in reality, the monitoring of the device state and the detection of the change point are continuously performed even after the series of operations is completed.

FIG. 9 is a sequence of fault information transfer according to the conventional fault information transfer method. In order to simplify the illustration,
The information transfer is illustrated as being performed between one NMS, one master CPU, and one slave CPU.

When the change point is detected by the slave CPU, the report is created as described above. For now, assume that the number of reports is four. The slave CPU sequentially transfers the created reports. First, the first report # 1 is transferred to the main CPU. The main CPU transfers the received report # 1 to the NMS.

The slave CPU then transfers report # 2 to the main CPU. Since the main CPU received the command from the NMS side at about the same time as this, the main CPU transfers the command to the slave CPU and transfers the report # 2 to the NMS. In this way, the main CPU plays a role of transferring the transferred data or command to the proper destination.

Since the slave CPU transfers the report unilaterally when detecting the change point, even if the command from the NMS side is received, the report is not transferred and the transfer of the report is continued. Since the transfer up to the report # 2 is completed now, the report # 3 and the report # 4 are sequentially transferred to the main CPU. Main CPU receives Report # 3 and Report # 4
To the NMS.

When the report # 4 is transferred, the slave C
The PU responds to the command. The main CPU that receives the response transfers it to the NMS. This completes a series of operations, and when the slave CPU next detects a change point, the report is transferred in the same sequence as above.

[0032]

As described above, the slave CPU
When it detects a change point, will unilaterally transfer the report to the main CPU.

Therefore, as shown in FIG. 8, even if a command is issued from the NMS side, the response is not immediately transferred to it. Since the command from the NMS side is transferred with the intention that the maintenance person on the NMS side confirms the current state of the device state, it is inconvenient that the slave CPU does not respond immediately.

Further, since the slave CPU unilaterally transfers the report to the main CPU, it is wasteful to continue the transfer even when the main CPU is busy or in an abnormal state and cannot receive the report. There is a risk of becoming.

In view of such a problem, the present invention provides a failure information transfer method capable of responding to a command even while a report is being transferred, and avoiding unnecessary transfer when the report receiving side cannot receive the command. The purpose is to do.

[0036]

The first means of the present invention is as follows:
This is a technique in which the slave CPU checks whether or not a command is received each time a report is transferred, and when a command is being received, the response to the received command is transferred before the next report is transferred.

According to the first means of the present invention, when a command is received, a response to the command is sent and then the next report is transferred. Therefore, the response to the command is delayed due to the transfer of the report. Disappear.

The second means of the present invention is that the slave CPU sends a report number to the report to be transferred, transfers the report, activates the primary timer, and receives the report.
Then, the confirmation information (ACK) is transferred with the report number attached, and the slave CPU clears the primary timer and transfers the next report when the received ACK number and the previously transferred report number match. In addition, it is a technique that stops the detection of the change point when the primary timer times out.

According to the second means of the present invention, since the slave CPU of the report can transfer the next report after confirming that the transferred report is correctly received, the main CPU is busy. Thus, there is no wasteful operation of transferring even if it cannot be received. Also, the fact that the primary timer times out means that the main CPU is busy for a long time or is in an abnormal state, so that the contents of the change point detection table cannot be transferred. At this time, since the change point detection table must be retained, it cannot be stored even if a new change point is detected. That is, by stopping the detection of the change point, it is possible to avoid the useless operation of detecting the change point information that cannot be stored in the change point detection table.

The third means of the present invention is that the slave CPU transfers the report to be transferred with a report number and starts the primary timer, and the main CPU adds the report number and sends confirmation information (ACK). When the slave CPU transfers the received ACK number and the previously transferred report number, the primary timer is cleared and the next report is transferred. When the primary timer times out, detection of the change point is stopped. In the second means of the present invention, when the primary timer times out and detection of the change point is stopped, the secondary timer is started to transfer all the failure information, and the secondary timer is received before all ACKs are received. This is a technique that adds a technique to prohibit the report when is timed out.

According to the third means of the present invention, since the report is prohibited by the timeout of the secondary timer, the unnecessary operation of transferring the unnecessary report when the main CPU is not ready to receive the report is avoided. It becomes possible to do.

The fourth means of the present invention is that the main CPU is activated when the main CPU is started up and when it is restored after being stopped in an abnormal state.
This is a technique for transferring a report start command from the PU to the slave CPU.

According to the fourth means of the present invention, after the main CPU is stopped for some reason and the sub CPU is in the report prohibition state, the sub CPU sends a report start command when the main CPU starts operation. Since it is transferred to the CPU,
The slave CPU can restart the report.

[0044]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a processing procedure in a slave CPU according to a first embodiment of the present invention.

The procedure of the first embodiment of the present invention will be described below with reference to the reference numerals in FIG. S1. Monitor the device status and detect changes. (Here, it is assumed that the change point is detected.) S2. Create a report. The number of reports is N (N is a positive integer).

S3. Set the counter to 1. S4. It is determined whether the count value of the counter is larger than the number N of reports. S5. When the count value of the counter is not larger than N (No), it is determined whether the command is received.

S6. When it is determined in step S5 that the command is received (Yes), the response to the command is transferred and the process proceeds to step S7.

S7. When it is determined that the command is not received in step S5 (No), step S5
In 6, the report is transferred after the response is transferred. S8. The count value of the counter is incremented by 1 and the process returns to S4.

Then, steps S4, S5, S6, S7 and S8 are performed until the count value of the counter becomes larger than N.
Is repeatedly executed. Then, when it is determined in step S4 that the count value of the counter is larger than N (Yes
In s), a series of operations is completed.

Since FIG. 1 shows the operation corresponding to the detection of one change point in a closed manner, a series of operations are supposed to be completed if the report is transferred in response to the command. Actually, even after the series of operations is completed, the monitoring of the device state and the detection of the change point are continuously performed. Therefore, in effect, after a series of operations is completed, it stands by until the next change point is detected.

As described above, in the first embodiment of the present invention, the presence / absence of a command is prioritized and determined, and if the command is received, the response transfer is prioritized over the report transfer. Therefore, the command from the NMS side will not be delayed due to the transfer of the report.

FIG. 2 is a sequence of fault information transfer according to the first embodiment of the present invention, showing a case where a command is received during the transfer of a report. The slave CPU detects a change point and creates a report. In this case, the number of reports shall be four.

The slave CPU transfers the report # 1 to the main CPU, and the main CPU transfers the received report # 1 to the NMS. Subsequently, the slave CPU transfers report # 2 to the master CPU.

At this time, the main CPU sends the report # 2 to the NMS.
Received a command from NMS before forwarding to. Therefore, the main CPU preferentially transfers the command received from the NMS to the sub CPU, and then transfers the report # 2 to the NMS.

Upon receiving the command transfer from the main CPU, the sub CPU preferentially transfers the response to the received command. When the main CPU receives the response from the sub CPU, the response is preferentially transferred to the NMS. Therefore, the NMS will transfer the command and then
The response can be received after the processing time of the PU and the delay time required for the transfer.

After that, the slave CPU transfers report # 3 and report # 4 to the main CPU, and the main CPU transfers report # 3 and report # 4 to the NMS. With that, the series of operations according to the first embodiment of the present invention is completed.

FIG. 3 shows a procedure of processing in the slave CPU according to the second embodiment of the present invention. The procedure of the second embodiment of the present invention will be described below with reference to the reference numerals in FIG.

S1. Monitor the device status and detect changes. (Here, it is assumed that the change point is detected.) S2. Create a report. The number of reports is N (N is a positive integer).

S3. Set the counter to 1. S4. It is determined whether the count value of the counter is larger than the number N of reports. S9. When the count value of the counter is not larger than the number N of reports (No), the primary timer is activated and the report with the report number is transferred to the main CPU.

S10. After transferring the report, main CPU
Then, it is determined whether or not the confirmation signal (ACK) having the same report number as the report transferred with the report number is received.

S11. When the ACK with the same report number as the report number of the report just transferred is received in step S10 (Yes), the primary timer is cleared.

S8. Next, the count value of the counter is incremented by 1, and the process returns to step S4. Until the count value of the counter becomes larger than the number N of reports, step S4,
S9, S10, S11 and S8 are repeatedly executed.
Then, when the count value of the counter becomes larger than the number N of reports (Yes in step S4), a series of operations ends.

S12. While the ACK having the same report number as the report number of the report just transferred is not received in step S10 (No), the operation of the primary timer is continued and it is determined whether or not the time is out. If the primary timer has not timed out (No), the process returns to step S8.

S13. If it is determined in step S12 that the primary timer has timed out (Yes), it is determined that the main CPU is not in the receivable state, and therefore the contents of the change point detection table must be retained. Therefore, the change point detection operation is stopped and a series of operations is ended.

Since FIG. 2 shows the operation corresponding to the detection of one change point in a closed manner, a series of operations are supposed to be completed when the report is transferred. In this case, Is effectively on standby until the next change point is detected.

According to the second embodiment of the present invention, the following two advantages occur. That is, since the transfer of the next report is permitted after receiving the ACK from the main CPU, it is not necessary to perform unnecessary transfer when the main CPU is busy.

When the primary timer times out, the main CPU cannot receive the change point information for a long time. At this time, the change point information must be held, and it is useless to detect the change point during this, but it can be avoided.

FIG. 4 is a sequence of fault information transfer according to the second embodiment of the present invention, and here shows a case where report transfer is completed smoothly. The slave CPU detects a change point and creates a report. Here again, it is assumed that the number of reports created is four.

The slave CPU activates the primary timer and transfers the report # 1 to which the report number is attached to the main CPU. The main CPU transfers the received report # 1 to the NMS and attaches to the received report # 1. The ACK is transferred to the slave CPU with the attached report number.

The slave CPU clears the primary timer when it receives an ACK with the same report number as that attached to the report # 1 just transferred. Next, sub CPU
Restarts the primary timer and transfers the report # 2 with the report number to the main CPU. The main CPU transfers the received report # 2 to the NMS and the report number attached to the received report # 2. Attach and ACK
To the slave CPU.

The slave CPU clears the primary timer when it receives an ACK with the same report number as that attached to the report # 2 just transferred. Thereafter, the same operation is continued to transfer all the reports created by the slave CPU to the master CPU, and the master CPU transfers all the reports received from the slave CPU to the NMS.

By doing so, the primary timer can be cleared and the next report can be transferred when the ACK with the same report number as the one attached to the report just received is received. . Therefore, it is possible to avoid the wasteful operation in which the CPU transfers the report when the main CPU is not in the receivable state.

Although not shown in FIG. 4, if the primary timer times out while an ACK with the same report number as the one transferred to it cannot be received, a change point is detected. So stop the main C
It is possible to avoid the useless operation of detecting the change point when the PU is not in the receivable state.

FIG. 5 shows a processing procedure in the slave CPU according to the third embodiment of the present invention. The procedure of the third embodiment of the present invention will be described below with reference to the reference numerals shown in FIG.

S1. Monitor the device status and detect changes. (Here, it is assumed that the change point is detected.) S2. Create a report. The number of reports is N (N is a positive integer).

S3. Set the counter to 1. S4. It is determined whether the count value of the counter is larger than the number N of reports. S9. When the count value of the counter is not larger than the number N of reports (No), the primary timer is activated and the report with the report number is transferred to the main CPU.

S10. After transferring the report, main CPU
Then, it is determined whether or not the confirmation signal (ACK) having the same report number as the report transferred with the report number is received.

S11. When the ACK with the same report number as the report number of the report just transferred is received in step S10 (Yes), the primary timer is cleared.

S8. Next, the count value of the counter is incremented by 1, and the process returns to step S4. Until the count value of the counter becomes larger than the number N of reports, step S4,
S9, S10, S11 and S8 are repeatedly executed.
Then, when the count value of the counter becomes larger than the number N of reports (Yes in step S4), a series of operations ends.

S12. While the ACK having the same report number as the report number of the report just transferred is not received in step S10 (No), the operation of the primary timer is continued and it is determined whether or not the time is out. If the primary timer has not timed out (No), the process returns to step S8.

S13. If it is determined in step S12 that the primary timer has timed out (Yes), it is determined that the main CPU is not in the receivable state, and therefore the contents of the change point detection table must be retained. Therefore, the change point detection operation is stopped.

S14. When the detection of the change point is stopped in step S13, the secondary timer is started. S15. After starting the secondary timer, move to the operation to transfer all reports. Here, step S15 is expressed as if it were a single step, but the operation of transferring the reports one by one is expressed by a single step.

S16. A for all transferred reports
Determine if CK was received. S17. When all ACKs have not been received in step S16 (No), it is determined whether the secondary timer has timed out. If the secondary timer has not timed out (No), the process returns to S15. S18. If the secondary timer times out in step S17 (Yes), the transfer of the report is prohibited and the series of operations is ended.

S19. All A in step S16
When it is determined that CK has been received (Yes
In s), the detection of the change point stopped in step S13 is restarted, and a series of operations ends.

As described above, by providing the ACK sequence, the slave CPU can know whether the master CPU is busy or not and transfer the report. Also, A
When CK is not returned for a predetermined time (when the primary timer times out), it is necessary to hold the contents of the change point detection table. Therefore, useless operation can be avoided by stopping the change point detection.

Further, since the secondary timer times out, it is judged that the main CPU side is in an abnormal state and cannot receive the report. Therefore, after that, the transfer of the report is prohibited to avoid the unnecessary operation. can do.

FIG. 6 is a sequence of fault information transfer according to the third embodiment of the present invention, in which after the primary timer times out, the secondary timer is activated to transfer the rest of the report and complete the transfer. Indicates.

The slave CPU detects a change point and creates a report. Here, it is assumed that the number of created reports is six. The secondary CPU activates the primary timer and transfers the report # 1 with the report number to the main CPU, and the main CPU transfers the received report # 1 to the NMS and the report number attached to the received report # 1. ACK is transferred to the slave CPU.

The slave CPU clears the primary timer when it receives an ACK with the same report number as that attached to the report # 1 just transferred. Next, sub CPU
Restarts the primary timer and transfers the report # 2 with the report number to the main CPU. The main CPU transfers the received report # 2 to the NMS and the report number attached to the received report # 2. Attach and ACK
To the slave CPU.

The slave CPU clears the primary timer when it receives an ACK with the same report number as that attached to the report # 2 just transferred. Similarly, the primary timer is started and the report # 3 is also transferred, and the AC with the report number attached to the transferred report # 3 is attached.
Receive K and clear the primary timer.

Further, the primary timer is activated, and the report # 4 is transferred with the report number, but from the main CPU, the AC with the report number added to the transferred report # 4 is attached.
Since K has not been transferred for a predetermined time, the primary timer times out.

Therefore, the secondary timer is started, and the report # 5 and the report # 6 to which the report numbers are attached are transferred one after another, and the main CPU waits for an ACK.

In the case of FIG. 6, ACKs corresponding to Report # 4 to Report # 6 are returned one after another. In the case of the secondary timer, the timer is not cleared until all ACKs are returned, so even if the ACKs corresponding to report # 4 and report # 5 are returned, the timer is not cleared, and the last report # The secondary timer is cleared only when an ACK corresponding to 6 is returned.

In the example shown in FIG. 6, since the ACK corresponding to the last report # 6 is returned before the secondary timer times out, the secondary timer is cleared, and the change check stopped when the primary timer times out The operation is restarted and the series of operations is completed.

If the secondary timer times out, the report is prohibited and the series of operations ends. However,
Even if the report transfer is prohibited, the monitoring of the device status is continued, and the latest device status is stored in the current value table.

As described above, when the secondary timer times out, is the main CPU in an abnormal state unable to receive data?
Although not in an abnormal state, the main CPU has not started up, so reception is not possible. At this time, it is useless to transfer the report, so it is necessary to prohibit the report. However, when the main CPU recovers from the abnormal state or when the main CPU starts up, that is, when the main CPU is stopped. It must be possible to transfer the report when it is released.

The fourth embodiment of the present invention is a form in which a technique of transferring a report start command from the main CPU is added so that a report can be transferred when the main CPU is released from the stopped state. .

Although not shown, in the procedure of FIGS. 1 and 3, a step of determining whether or not a report start command is received from the main CPU is inserted before step S1, and a step is executed when the report start command is received. Move to S1,
If the report start command is not received, the loop may be repeated to repeat the above determination.

In the procedure of FIG. 5, a step of determining whether or not a report start command from the main CPU is received is inserted before step S1, and when the report start command is received, the process proceeds to step S1 and the report is sent. If a start command has not been received, a loop is made to repeat the above determination, and a step for determining whether or not a report start command has been received from the main CPU is inserted immediately after step S18, and the report start command is set. If a report start command is not received, a series of operations may be ended when it is received, and the above determination may be repeated by looping.

According to the fourth means of the present invention, it is possible to cause the slave CPU, which is once prohibited from transferring the report, to transfer the report again. FIG. 7 is a sequence of fault information transfer according to the fourth embodiment of the present invention, in which a technique of transferring a report start command from the main CPU when the main CPU is released from the stopped state is added to the procedure of FIG. The sequence in the case of doing is shown.

That is, when the main CPU starts up, the report start command is transferred to the sub CPU. Upon receiving the report start command, the slave CPU starts change point detection and creates a report.

After that, the report is transferred in the sequence already described. Then, it is assumed that after the report # 2 is transferred, the main CPU falls into an abnormal state and cannot be received.

In the slave CPU, the AC corresponding to the report # 2
Since K is not returned, the primary timer times out. Therefore, the secondary timer is activated, the remaining reports are transferred, and the ACK is waited for.

However, since the ACK is not returned, the secondary timer times out, and it is determined that the main CPU is abnormal. Therefore, the report is prohibited and the latest device status is continuously monitored. As it starts and transfers the report start command, it is used as a trigger for the slave CPU
Restarts the detection of the change point, creates a report and transfers it to the reception signal CPU. After that, the sequence operates as described above.

The above consistently shows that the main C is on the main shelf.
The case where the PU and the sub CPU are provided has been described. This is because when the number of devices to be monitored is large or the number of monitoring points in the devices to be monitored is large, it is possible to add and distribute the CPU for monitoring the device state and the CPU for transferring the change point information. Further, if the functions are shared as described above, there is an advantage that the system can be flexibly dealt with even when the monitoring target is expanded in the future.

However, when the number of devices to be monitored is small, the number of monitoring points to be monitored is small, or when future expansion is not expected, a single CPU is used to monitor the device status and transfer change point information. It is also possible to do. In this case,
When a single CPU monitors the device status and transfers change point information, the above problem becomes a problem between the CPU and the NMS.

Therefore, considering that the unit CPU monitors the device status and transfers the change point information, the slave CPU in the above is the CPU that transfers failure information, the main CP.
U should be regarded as the CPU that receives the transfer of fault information.

[0108]

As described in detail above, according to the present invention, the response to the command is not delayed due to the transfer of the fault information. Further, it becomes possible to transfer the fault information by knowing whether the side receiving the fault information transfer is busy, and it is possible to stop reporting when the side receiving the fault information transfer is in an abnormal state. It is possible to avoid transfer of various fault information. In addition, the failure information can be automatically transferred when the side receiving the failure information has been started up or when the abnormal state is exited.

[Brief description of drawings]

FIG. 1 shows a first embodiment of the present invention.

FIG. 2 is a sequence of fault information transfer according to the first embodiment of the present invention.

FIG. 3 is a second embodiment of the present invention.

FIG. 4 is a sequence of fault information transfer according to the second embodiment of the present invention.

FIG. 5 is a third embodiment of the present invention.

FIG. 6 is a sequence of fault information transfer according to the third embodiment of the present invention.

FIG. 7 is a sequence of fault information transfer according to the fourth embodiment of the present invention.

FIG. 8 is a conventional transfer method.

FIG. 9 is a sequence of fault information transfer according to the conventional transfer method.

FIG. 10 is a configuration example of a system to which the invention is applied.

FIG. 11 shows the configuration of the CPU.

FIG. 12 is a diagram illustrating collection of device states and change point detection.

FIG. 13 is a command configuration example.

FIG. 14 is a configuration example of a response and a report.

[Explanation of symbols]

 1 NMS 2a, 2b Main shelf 3a, 3b Secondary shelf 4 Communication path between NMS and primary shelf 5 Communication path between primary shelf and secondary shelf

Claims (5)

[Claims] 1. When a sender of fault information receives a command from a receiver of fault information while transferring a report of fault information, it transfers a response to the command and then transfers the remaining report of fault information. A fault information transfer method characterized by the following. 2. The fault information sender assigns a number to the report of the fault information, transfers each report by activating the first timer, and the fault information receiver attaches the received report to the received report. The confirmation information to the sender of the failure information, and the sender of the failure information confirms that the received confirmation information is valid confirmation information with a number that matches the number of the previously transferred report. In some cases, the first timer is cleared to allow the transfer of the next report, and when valid timer information cannot be received for a predetermined time and the first timer times out, the failure information to be transferred is displayed. A fault information transfer method characterized by stopping generation. 3. The fault information transfer method according to claim 2, wherein when the first timer times out, a second timer is activated to transfer a report of remaining fault information, and the report for all the reports is transmitted. A failure information transfer method, wherein transfer of a failure information report is prohibited when valid confirmation information cannot be received for a predetermined time and the second timer times out. 4. The fault information transfer method according to claim 2, wherein when the transmitter of the fault information receives a command from the receiver of the fault information while transferring a report of the fault information, the fault information is transferred to the command. A failure information transfer method characterized by transferring the remaining failure information report after transferring the response. 5. The fault information transfer method according to claim 1, wherein the fault information receiving side reports fault information to a fault information transmitting side when the fault information receiving side is in a receivable state. A failure information transfer method, in which a command requesting transfer of the failure information is transferred, and when the sender of the failure information receives a command requesting transfer of the report of the failure information, generation of the failure information is started.