JPS6161141B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a host processing device, a data processing control device,
The present invention also relates to a failure reporting method from the lower-level device to the higher-level processing device via the data processing control device in a data processing device having a lower-level device.

Conventionally, data processing control devices operate by sending a number of second commands to lower-level devices in response to a first command from a higher-level processing device, and have to retry several times in response to a failure that occurs during that time. Alternatively, a retry using an alternate path is performed to recover from the failure or isolate the failed part. During this time, each time the data processing control device receives a fault report sent from a lower-level device or detects a fault, the data processing control device
Since fault information is created and reported to the higher-level processing device, continued operation of the data processing control device is forced to wait until the higher-level processing device receives the fault information.

As described above, if there are many intermediate failure reports during operation, the number of interrupts to the host processing device and the waiting time of the data processing control device increase, resulting in a decrease in overall processing efficiency.

Note that the upper-level processing device that should perform intermediate failure reporting may be the upper-level processing device that issued the above-mentioned first command, or a system that is different from the upper-level processing device that issued the above-mentioned first command. In some cases, it is a supervising higher-level processing device.

The present invention stores intermediate failure information that occurs in the data processing control device during the operation of the data processing control device in response to the first command from the host processing device, and After all operations are completed, the saved intermediate failure information is reported to the higher-level processing unit that should report the failure, thereby reducing interruptions to the higher-level processing unit that should be reporting the failure, and improving the data processing control unit. The object of the present invention is to shorten the operation time for the above-mentioned first command.

According to the present invention, when a first command sent from a higher-level processing device is received, a second command necessary for executing the first command is sent to a lower-order device, and the second command is sent from the lower-order device. data that sends a second report that the first command has been correctly executed to the higher-level processing device upon receiving a first report that the second command information that is sent as response information to the command has been correctly executed; The processing control device is provided with a fault information storage means and a counter indicating the number of fault information stored in the fault information storage means, and the data processing control device performs the above-mentioned processing in response to the sent out second command. Every time a failure report is received as the response information from a lower device, or when neither the first report nor the failure report is received as the response information, the failure information is created and the failure information is used as the failure report. causes the information storage means to sequentially store the information, causes the counter to count by one, resends the second command, and upon receiving the first report, indicates that the failure information is stored in the failure information storage means. A fault handling method is obtained in which, if the counter is displaying, the fault information stored in the fault information storing means is sent to the higher-level processing device to which the fault report should be made.

Next, an embodiment of the present invention will be described with reference to the drawings while comparing it with a conventional system.

FIG. 1 shows a conventional method, and FIG. 2 shows a method according to the present invention. In both FIG. 1 and FIG. 2, in response to the first command 10 sent from the higher-level processing device 1, the data processing control device 3 further sends several instructions to the lower-level device 4 in order to execute the first command 10. A second command is issued, an operation is performed, and after the operation is completed, a completion report 11 for the first command 10 is sent to the higher-level processing device 1. As an example of the processing method when a failure occurs during this period, the second command from the data processing control device 3 to the lower-level device 4 is successfully executed after three retries, and the failure report during operation is Figures 1 and 2 show the case where the
This will be explained with reference to the diagram. In the conventional system shown in FIG. 1, when a fault report 24 is received in response to the second command 20 issued from the data processing control device 3 to the lower-level device 4, the data processing control device 3 sends the fault report 24 to the faulty device. and information indicating that it is an intermediate failure report is added to create failure information, and the system interrupts the upper processing unit 2 that controls the system to report the failure information 30, which is accepted. Wait until then without performing any other processing. When the signal 33 indicates that the failure information 30 has been accepted, the data processing control device 3 issues the second command 21 for retry.
is sent to the lower-level device 4 again. At this time, when the failure report 25 is reported again, the data processing control device 3 adds additional information as in the case of the failure report 24, interrupts the system supervising higher-level processing device 2, and reports the failure information 31. Acceptance completion signal 3 for it
4, the data processing control device 3 sends the second command 22 to the lower-level device 4 for a second retry. The corresponding failure report 26 is also reported as failure information 32 to the system-intensive upper-level processing unit 2 as before, and when acceptance is indicated by the signal 35, a second command 23 for a third retry is sent. . This command 23 is correctly executed in the lower device 4 and the final report 27
is reported, the data processing control device 3
sends a completion report 11 to the higher-level processing device 1 that issued the first command 10, and the operation is completed.

Next, referring to FIG. 2, the operation of the present invention will be explained. Similar to FIG. 1, the first
When the data processing control device 3 receives the command 10,
A second command 20 is sent to the lower-level device 4 to perform the operation.
will be sent. On the other hand, when a failure report 24 is sent from a lower-level device, the data processing control device 3 adds information indicating that it is a failed device and an intermediate failure report to the failure report 24 to create failure information. The failure information is stored in a buffer 5 in the data processing control device 3, and a counter 6 indicating the number of failure information is incremented. Thereafter, a second command 21 is issued for retry. On the other hand, the failure report 25 that is reported again is also stored in the buffer 5 with additional information attached, and the failure information counter 6 is incremented. Failure report 26 for second command 22 for second retry
The same is true. When the lower-level device 4 operates normally according to the second command 23 for the third retry, and the operation according to the first command 10 from the upper-level processing device 1 is completed according to the completion report 27, the data processing control device 3 , sends a completion report 11 to the higher-level processing device 1. Thereafter, the data processing control device 3 determines whether or not intermediate fault information has been received during the previous operation based on the contents of the counter 6, and if fault information has been received, an interrupt is sent to the system supervisory host device 2. , of the information in the buffer 5, the counter 6
Only the number of failure information 40 shown in is reported.

Note that in the above explanation and in Figures 1 and 2, the higher-level processing device 1 that issues commands and the higher-level processing device 2 that unifies the system are separate processing devices, but these two higher-level processing devices are the same. Even in that case, there is no change in the operation of the data processing control device. In the embodiment of the present invention shown in FIG. 2, the data processing control device 3 issues a second command to the lower-order device 4, but the lower-order device 3 sends a report 27 indicating that the command was correctly executed as response information. Trouble report 24, 25, 2
Even if none of 6 is received, a failure is detected and failure information is created in the same way as above,
The failure information is stored in the buffer 5, the counter 6 is counted by one, and the second command is re-sent to the lower-order device 4.

As described above, according to the present invention, the data processing control device stores failure information during operation in response to the first command from the host processing device in the failure information storage means, and after completing the operation in response to the first command. If there is any fault information in the fault information storage means, it is configured to report it to the higher-level processing device that should report the fault, thereby reducing the number of interrupts to the higher-level processing device that should report the fault, and reducing the number of interruptions to the above-mentioned data processing control device. This has the effect of increasing the speed of the operation in response to the first command.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a conventional fault handling method, and FIG. 2 is a diagram showing a fault handling method according to an embodiment of the present invention. In both figures, 1 is the command issuing upper processing unit, 2
3 is a system-integrating upper processing device, 3 is a data processing control device, and 4 is a lower device. Further, 10 is a first command from the upper processing device, 11 is a completion report from the data processing control device in response to the first command, and 2
0, 21, 22, 23 are second commands from the data processing control device to the lower device, 24, 25, 26,
27 shows the response of the lower-level device to the second command. Also, in Figure 1, 30, 31, 32
is a fault report from the data processing control device to the system-integrating upper-level processing device; 33, 34, and 35 are acceptance signals therefor and signals for continuing the operation of the data processing control device thereafter. Also second
In the figure, 5 is a buffer for storing intermediate failure information, and 6 is a counter indicating the number of failure information. Furthermore, 40 indicates the transfer of information in the buffer to be sent to the system-integrating upper-level processing unit when failure information occurs during command operation.

Claims (1)

[Claims] 1 Upon receiving the first command sent from the upper processing device, sends the second command necessary for executing the first command to the lower order device, and sends it as response information to the second command from the lower order device. a data processing control device that, upon receiving a first report that the second command information is correctly executed, sends a second report that the first command has been correctly executed to the higher-level processing device; A fault information storage means and a counter indicating the number of fault information stored in the fault information storage means are provided, and the data processing control device receives the response information from the lower-order device in response to the sent second command. Each time a failure report is received as response information or neither the first report nor the failure report is received as response information, the failure information is created and the failure information is sequentially stored in the failure information storage means. and causes the counter to count by one, and resends the second command, and
If the counter indicates that the fault information is stored in the fault information storage means, the higher-level processing should report the fault information stored in the fault information storage means. A fault handling method characterized by sending data to a device.