JPS62202249A - Shared resource access guard system among plural host computers - Google Patents

Abstract

PURPOSE:To switch shared resources quickly and to sufficiently guard shared resources against access as well by allowing each host to generate closing information based on a logical host number inputted to the host itself and shared resource information and to check this generated information to decide whether access is permitted or not. CONSTITUTION:A logical host number input means 5 inputs a logical host number. An information input means 7 reads in shared resource definition information stored in a shared resource definition information storage area 13. A closing means 8 generates the closing information of shared resources, which the host itself cannot access, in accordance with the logical host number from the means 5 and shared resource definition information from the means 7 and registers this generated information in a closing information storage area 12. It is checked whether this registration is terminated normally or not, and the processing is terminated if it is terminated normally, but the error processing is performed if it is terminated abnormally.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a system that includes two or more host computers (hereinafter simply referred to as hosts) and shared resources that can be shared by each host. This invention relates to a method for guarding access from posts other than a predetermined host.

[Conventional technology]

Conventionally, the following two methods have been used to guard access to shared resources among multiple hosts.

(A) A method in which shared resources that can be shared by multiple hosts, except those used by a certain host, are physically separated from that host.

In this method, when a shared resource such as another external storage device is allocated to the host, the shared resource must be physically reconnected. In terms of guarding access to shared resources, there is no problem since they are physically separated.

(B) Physically connect each host to all shared resources and allow each host to access all shared resources, for example, except during certain hours, only one host can use a shared resource. A method of operation that allows access only to the public.

[Problem that the invention seeks to solve]

Among the conventional methods mentioned above, method (A) does not have any problems in terms of access guard, but
When allocating shared resources used by one host to another host for some reason, it is necessary to physically switch the shared resources, which has the drawback of requiring a lot of time.

Therefore, in a system that consists of two or more hosts, at least one of which is on standby so that it can take over the processing of another active host, if the standby host goes down, the active The disadvantage is that it takes a long time to be able to take over the host's processing.

On the other hand, in method (B), there is no need to physically switch connections, so shared resources can be switched relatively quickly, but since all shared resources can be accessed, operation The drawback is that there is a high risk that an error will destroy the contents of shared resources allocated to other hosts.

The present invention solves these conventional drawbacks, and its purpose is to enable quick switching of shared resources and to sufficiently guard access to shared resources.

[Means for solving problems]

In order to achieve the above object, the present invention provides a system for guarding access from host computers other than a predetermined host computer to a shared resource that can be shared among a plurality of host computers, in which the plurality of host computers are connected to the shared resource. Each of the plurality of physically connected host computers has: a logical host number input means for inputting a logical host number when starting up the host; and a logical host number corresponding to a shared resource definition information storage area on an external storage medium. a blocking means that generates blocking information based on the shared resource definition information stored in the shared resource definition information and the manually entered logical host number, and stores the generated blocking information in a blocking information storage area; Blockage checking means is provided for checking blockage information stored in the blockage information storage area to determine whether or not access is permitted.

[Effect]

Each post generates blockage information based on the logical host number input to its own host and shared resource definition information, and when accessing, this blockage information is checked to determine whether or not access is possible. By determining the contents of shared resource definition information corresponding to each host according to the shared resources to be used by each host, it is possible to prevent erroneous access to shared resources of other hosts that are assigned different logical host numbers. In addition, when switching shared resources used by one host to another host, for example, it is possible to give the logical host number assigned to the switching source host to the switching destination host, and physical connection switching is possible. Since this is not necessary, shared resources can be switched relatively quickly.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of a host used in the present invention, including two or more hosts, at least one of which is a host.
1 shows an example of a host structure suitable for a system that is on standby as a standby host so that it can take over the processing of another active host, where 1 is a processing device and 2 is an external storage medium. Although the host in this embodiment is composed of a processing device 1 and an external storage medium 2, it is of course possible to adopt a configuration in which the external storage medium 2 is shared by a plurality of hosts. The external storage medium 2 has a shared resource definition information storage area 13, and this storage area 13 contains shared resource definition information in advance corresponding to the logical host number, that is, which resources among the resources that can be shared by multiple hosts are stored. Information regarding which hosts with logical host numbers can or must not be used is stored in advance. The processing device 1 also includes monitoring means 3. Processing continuation means 4. Logical host number takeover means6. Logical host number input means 5. Resource information input means 7. Closing means8. Blockage checking means 9. Blockage check bypass means IO1 Blockage release means 11. It includes a blockage information storage area 12.

The logical post number manual means 5 is a means that allows a logical host number to be externally set and input to this host, and an example of this process is shown in FIG.

The resource information input means 7 is a means for inputting the shared resource definition information stored in the shared resource definition information storage area 13 of the external storage medium 2, and an example of its processing is shown in FIG.

The blocking means 8 blocks access from its own host based on the logical host number manually entered by the logical host number input means 5 and the shared resource definition information inputted from the shared resource definition information storage area 13 by the resource information input means 7. A processing example is shown in FIG. 5.

The operation mode of the monitoring means 3 differs depending on whether the host is an active host or a standby host, and in the case of an active host, it monitors abnormalities in the host's operation and stoppages due to failures, and monitors standby posts. If so, it communicates with the active host and monitors the active host for operational abnormalities and stoppages due to failures. An example of processing on an active host is illustrated in Figure 6.
An example of processing at the standby host is illustrated in FIG.

The processing continuation means 4 operates effectively when this host is a standby host, and when another active post is stopped due to an abnormal operation or a failure, the processing that was being performed on the active host is switched to the standby host. An example of this process is shown in FIG.

The logical host number transfer means 6 also operates effectively when this host is a standby host, and is a means for inheriting the logical host number when processing from another active host is switched to the standby host, for example, as shown in FIG. Execute the process shown in

The blocking release means 11 operates effectively when this host is a standby host, and when the processing of the active host is switched to the standby host and the logical host number is also taken over, the blockage release means 11 is configured to release the blockage until then according to the logical host number. This is a means of releasing resources that have been previously used, and FIG. 10 shows an example of this process.

The blockage checking means 9 is a means for checking the blockage information stored in the blockage information storage area 12 to determine whether access is possible when the host accesses the shared resource. An example of its operation is shown in FIG. It is shown.

The blockage check bypass means 10 operates effectively when this host is a standby host. When switching from an active host to a standby host is necessary, in order to take over processing as quickly as possible, the standby host must prepare in advance by opening an access bus to the shared resources. To prepare for this, it is necessary to temporarily enable access to the shared resource that is already blocked at that time, and the blockage check bypass means 10 makes this possible. In addition, blockage check bypass means 1
0 executes the process shown in FIG. 12, for example.

In FIG. 2, three hosts having the configuration shown in FIG. 1 are used, of which two hosts (processing device A) are used.

An example of a system configuration is shown in which the host (including the processing device B) is set as the active host (21, 22), and the remaining host (the host including the processing device C) is the standby host 23. Each host 21-23 has a shared resource (R,) 27. which is accessible from them. (R,) Physically connected to 2B. In addition, the active host) 21.22 and the standby host 23 have logical host numbers rlJ and r2J, respectively.
, is running on r3J. Furthermore, each host can communicate with each other for monitoring purposes by means of monitoring means 24,25.26 (corresponding to the monitoring means 3 in FIG. 1) located therein.

Next, the operation of the embodiment of the present invention will be described in detail with reference to the drawings.

First, when a host is set up and taken down, a logical host number is input, and according to that number, shared resources that must not be accessed by that host are blocked from the information in the shared resource definition information storage area 13. The operation will be explained in detail with reference to FIGS. 3, 4, 5, and 6 while explaining the operation of each means in FIG. 1.

The logical host number input means 5 in FIG. 1 inputs identification information as to whether the own host is an active host or a standby host in step 31 in FIG. 3, and inputs the logical host number in the next step 32. The input information is checked for validity in step 33, and if it is valid, the process is terminated. By performing such processing for each host, the logical host number "1" is set to the active host 21 shown in FIG. 2, the logical host number "2" is set to the active host 22, and The logical host number “3” is set to the standby host 23.

Next, the resource information input means 7 in FIG. 1 reads the shared resource definition information stored in the shared resource definition information storage area 13 in step 41 in FIG. 4, and checks for read errors in step 42. If there are no errors, end the process,
If there is an error, error processing is performed and the process ends abnormally.

Next, in step 51 of FIG. 5, the blocking means 8 of FIG. Blockage information of the inaccessible shared resource is generated and registered in the blockage information storage area 12. Next, in step 52, it is checked whether the above registration has been completed normally. If it is normal, the process is ended, and if it is abnormal, error processing is performed in step 53. Now, as shown in FIG. 13(a), if the shared resource 27 in FIG. 2 is used only by the active host 21, and the shared resource 28 is used only by the active host 22, the above-mentioned processes will The shared resource 2 is in the blockage information storage area 12 in the host 21.
8 is set to prohibit access, the shared resource a27 is set to be access prohibited in the blockage information storage area 12 of the post 22, and the shared resources 27 and 28 are set in the blockage information storage area 12 of the standby host 23. Access will be prohibited for both parties.

Now, after the host has started up as described above, the monitoring means 3 etc. operate, and in the case of this embodiment, the active host 21
Alternatively, when the active host 22 goes down due to a failure or the like, the process is automatically executed by the standby host 23.The operation at that time will be described later, but first,
How each host performs access guarding of predetermined resources will be explained with reference to FIGS. 1, 11, and 12.

In the active host) 21 and 22, when an input/output command for accessing a shared resource or another resource is issued from a data management program or the like, the blockage checking means 9 is activated. When this blockage checking means 9 is activated, the first
In step 111 of FIG. 1, it is checked whether the blockage check bypass flag is set, and if it is set, access to the resource is unconditionally permitted. This blockage check bypass flag is provided to break the blockage and enable access to the shared resource as a preliminary preparation in the standby host, as will be described in detail later.

Conversely, if the blockage check bypass flag is not set, the blockage check required parameter in the input/output command is accepted in step 112, and blockage information regarding the corresponding resource is retrieved from the blockage information storage area 12 in the next step 113. Then, in the next step 114, it is determined whether or not the target blockage information exists. If it does not exist, access to the resource is permitted for this host, and the I/O command is notified that access is allowed. (Step 1
16).

On the other hand, if it exists, it is determined in step 115 whether or not actual access is possible using a blockage flag that indicates whether or not to implement blockage based on the blockage information, and if access is possible, a step 116 indicates that give notice of
If access is not possible, a notification to that effect is sent in step 117. The issued input/output command is executed if the notification from the blockage check means 9 indicates that access is possible, but if access is not permitted, the input/output command is not executed.

Therefore, access to shared resources that are blocked on each host is prohibited.

The above blockage check means 9 prohibits access to resources other than the shared resources corresponding to the preset logical host number, and in the case of the standby host 23 in FIG. 2, the shared resources 27, 28 cannot be accessed as is. However, if the active hosts 21 and 22 go down due to some kind of failure, in order for the standby host 23 to take over the processing as quickly as possible, the standby host must immediately access the shared resources after starting up. It is necessary to establish an access bus to the area. In this way, the blockage check bypass means 10 shown in FIG. The operation will be explained in detail with reference to FIGS. 11 and 12.

The blockage check bypass means 10 is a means that is allowed only in a specific section of a specific process, is activated prior to a request for access to a shared resource, and is a means for setting and canceling the blockage check bypass. When the occlusion check bypass means 10 is activated, step 12 in FIG.
In step 1, it is determined whether the request from the specific task is a request to start blockage check bypass processing, and if it is a request to start blockage check bypass processing, in step 122, a blockage check bypass flag is set and the blockage check bypass processing is started. If it is determined that the request is not a start request but an end request, the blockage check bypass flag is reset in step 123.

In this way, when the occlusion check bypass flag is sent by the occlusion check bypass means IO, the 11th
As explained in the figure, access to the shared resource will be permitted, and the standby boss 23 will be able to prepare for opening an access bus.

Now, next is the monitoring means 3 in FIG. Processing continuation means 4, logical host number takeover means 6. The operation when the processing of the active host is taken over by the processing of the standby host using the blocking release means 11 will be explained.

First, the operation of the monitoring means 3 will be explained in detail with reference to FIGS. 6 and 7.

The monitoring means 3 operates differently depending on whether the host is an active host or a standby host. In the case of an active host, the process shown in Figure 6 is performed, and in the case of a standby host, the process shown in Figure 7 is performed. Process.

In the case of an active host, an event is issued to a predetermined monitoring task in step 61 of FIG. 6, and step 62
and wait for the response of that event. Next, in step 63, if the answer is not returned within a certain period of time or is abnormal, the host down process is performed in step 65, and a down notification is issued to the standby host in step 66, and the process ends. . Further, if the response is normal in step 63, an event indicating that the own host is normal is issued to other hosts within a predetermined period in step 64.

On the other hand, when the monitoring means 3 in the standby host receives a down notification of another host in step 71 of FIG.
In step 72, a down notification is sent to the processing continuation means 4 in FIG. 1 to start it up.

Next, the operation of the standby host accepting a down notification and continuing processing will be explained in detail with reference to Figures 8, 9, and 10 while explaining the operation of each means in Figure 1. do.

When the processing continuation means 4 in FIG. 1 receives the down notification from the monitoring means 3 in step 81 in FIG. A notification is sent to the takeover means 6. The logical host number takeover means 6, which received this notification in step 91 of FIG. 9, performs logical host number takeover processing in step 92, and requests the resource information input means 7 of FIG. 1 to take in resource information in step 93. Issue. Upon receiving the request, the resource information input means 7 reads the resource definition information from the shared resource definition information storage area 13 in step 41 of FIG. 4, and activates the block release means 11.

When the blocking release means 11 in FIG. 1 is activated, in step 101 in FIG. According to the resource definition information, the blockage information that has already been blocked in the standby host is deleted, and in the next step 102, it is checked whether the deletion has been completed normally, and if it is normal, the process is terminated,
If the process ends abnormally, error processing is performed in step 103.

The logical host number handover means 6 in FIG. 1 confirms the resource information processing by the resource information manual means 7 and the blockage release means 11 as described above in step 94 in FIG. 9, and then sends a processing completion notification in step 95. The processing continuation means 4 performs the processing.

The processing continuation means 4, which has received the above notification at step 85 in FIG. In the next step 87, the processing jobs or tasks that have been waiting until then are notified of restarting the processing.

Through the above-described processing, if the active host 22 fails due to a failure or the like in FIG. 13(a), the first
As shown in FIG. 3(b), the standby host 23 in FIG. 13(a) takes over the logical host number "2" and becomes the active host 2.
When the temperature reaches 3°, access to the shared resource 28 used by the active host 22 becomes possible. In addition, the first
In Figure 3(b), the standby host 22', which was originally the active host 22, is given the logical host number "3",
Indicates a standby state as a new standby host. Thus, in the present invention, logical host numbers are inherited without regard to individual processing units or hosts, and shared resources are inherited according to logical host numbers.

〔Effect of the invention〕

As explained above, according to the present invention, each host blocks resources that should not be accessed from the host based on the logical post number input to the host and the shared resource definition information corresponding to the logical number. It generates information and checks this blockage information at the time of access to determine whether access is allowed or not, and has the effect of guarding against access from hosts other than the designated host to shared resources that can be shared among multiple hosts. .

Furthermore, since each shared resource is physically connected, there is an advantage that the shared resource can be instantly switched from one host to another.

Furthermore, since shared resource definition information is provided corresponding to logical host numbers independent of physical hosts, for example, if the present invention is configured with two or more hosts, at least one of which is used as a standby host, When applied to a system that is on standby so that it can take over the processing of another active host, there is no need to decide on a specific host as the standby host, and it has the advantage that it can be changed sequentially.

[Brief explanation of drawings]

Fig. 1 is a block diagram of an embodiment of the host used in the present invention, Fig. 2 is a block diagram showing an example of a system consisting of multiple hosts having the configuration shown in Fig. 1, and Fig. 3 is a logical host number input. 4 is a flowchart of an example of processing of the resource information input means 7, FIG. 5 is a flowchart of an example of processing of the blocking means 8, and FIG. 6 is a flowchart of an example of processing of the monitoring means 3 in the active mode. FIG. 7 is a flowchart of a processing example of the monitoring means 3 in standby mode, FIG. 8 is a flowchart of a processing example of the processing continuation means 4, and FIG. 9 is a flowchart of a processing example of the logical host number takeover means 6. Flowchart, FIG. 10 is a flowchart of a processing example of the blockage release means 11,
12 is a flowchart of an example of processing by the blockage check means 9, FIG. 12 is a flowchart of an example of processing by the blockage check bypass means 10, and FIG. 13 is a diagram showing an example of the system state before and after the host goes down. In the figure, 1...processing device, 2...external storage medium, 3...monitoring means, 4...processing continuation means, 5...
Logical host number input means, 6... Logical host number transfer means, 7... Resource information input means, 8... Blockage means, 9... Blockage check means, 10... Blockage check bypass means, 11 . . . Block release means, 12 . . . Blockage information storage area, 13 . . . Shared resource definition information storage area.

Claims (1)

[Claims] In a method for guarding access from host computers other than a predetermined host computer to a shared resource that can be shared among a plurality of host computers, the plurality of host computers are physically connected to the shared resource. , and each of the plurality of host computers includes: a logical host number input means for inputting a logical host number at the time of host startup; and a shared resource definition information storage area on an external storage medium stored in correspondence with the logical host number. A blocking means that generates blocking information based on the resource definition information and the input logical host number and stores the generated blocking information in a blocking information storage area; 1. A shared resource access guarding method between a plurality of host computers, comprising blockage check means for checking blockage information stored in a computer to determine whether or not access is permitted.