JP2001147888A - Method for recognizing connecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To avoid a situation in which it is impossible to detect a device, and to attain high speed device search in a server system equipped with a plurality of servers 110 and 120 and a shared disk 100, where each server is provided with boot information and the shared disk is capable of defining a plurality of device numbers and logical unit numbers. SOLUTION: Respective servers 110 and 120 preliminarily register device numbers and logical unit numbers to be accessed in boot information, and at the time of starting the system, the server searches only the device and logical unit whose numbers are registered, and allows its own system to recognize them.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a server system comprising a plurality of servers and a disk storage device (shared disk) shared by the servers.
The present invention relates to a connected device recognition method capable of searching for only a device or a logical unit to be accessed.

[0002]

2. Description of the Related Art FIG. 3 shows a process of creating a device table for storing and managing information on connected devices in a conventional method for recognizing connected devices in a server system as described above.
It will be described based on. As shown in FIG.
I ID = 0, LUN (logical unit number) = 0
Is set in the system (step 30).
1).

Next, a test unit ready (Test) is used to check whether a valid device is connected or not.
A Unit Ready) command is issued (step 302). If a valid device is not connected to the address to which the command has been issued, a selection timeout occurs, so it is checked whether a selection timeout has occurred (step 303).

If no selection timeout has occurred (NO in step 303), that is, if there is a response from the device, a device table for storing and managing information on connected devices is created (step 304).

After the device table is created, the LUN is checked in order to check whether the device is connected to the next LUN.
Is counted up (step 305), and it is determined whether or not the LUN exceeds a given value (step 3).
06).

If the value does not exceed the value (NO in step 306), the process returns to the step of issuing a test unit ready command (step 302), and the process is repeated. On the other hand, if it exceeds the value (step 30
6 is YES) and when a selection timeout occurs in step 303 (step 303 is Y
In the case of ES, LUN is reset to 0 (step 307).

Next, the SCSI ID is counted up (step 308), and thereafter, it is determined whether or not the SCSI ID exceeds a value that can be given (step 30).
9). If the value exceeds the value (YES in step 309), the process ends, and if the value does not exceed the value (step 3).
If 09 is NO), the process returns to the step of issuing a test unit ready command (step 302). Thereafter, in step 309, until the value exceeds (Y
ES), and steps 302 to 308 are repeated.

As can be seen from the above-described device table creation process, the conventional method of recognizing a connected device in a shared disk type server system is based on a search from the smallest device number to the upper limit of the number.

[0009]

The above prior art has the following problems. First, in an environment in which a young number is not assigned to a device, there is a problem that a device to be recognized cannot be detected.
This is because if the device has a two-stage device number structure such as an ID and a LUN in SCSI, and if it is determined that the device is not connected by a device search with a smaller LUN, the process is performed in order to increase the processing efficiency. The next search is not performed and the next SCS
This is because the process has shifted to the search process for the ID.

Second, there is a problem that it takes time to search for devices. This is because it is necessary to detect a selection timeout in order to recognize that the device is not connected, but it is unknown how far the device is connected (in the case of SCSI, detection processing is required for several IDs).
This is because the detection process must be performed for a certain number of units.

An object of the present invention is to provide a method for recognizing a connected device that does not cause a situation in which a device cannot be detected and that can perform a high-speed device search.

[0012]

SUMMARY OF THE INVENTION The object is to provide a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device can define a plurality of device numbers. In a server system, a connection device characterized in that each server registers in advance the number of a device to be accessed by the server in the boot information, and searches for only the registered device at the time of system startup and causes the system to recognize the device. Achieved by recognition methods.

The above object is also achieved in a server system comprising a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device being capable of defining a plurality of logical unit numbers. Connected device recognition in which each server registers in advance the logical unit number that the server wants to access in the boot information, and when starting up the system, searches only the logical unit whose logical unit number has been registered and recognizes it in its own system. Achieved by the method.

Further, the above object is to provide a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device defining a plurality of device numbers and a plurality of logical unit numbers. In a server system that can be used, the server number and the logical unit number that each server wants to access are registered in the boot information in advance, and when the system starts up, only the device and the logical unit whose number and logical unit number have been registered are searched. This is achieved by a connected device recognition method characterized by causing its own system to recognize.

Further, the above object is achieved by a connection device recognition method according to claim 1 or 3, wherein the device is a SCSI device.

Further, the above object is attained by a connection device recognition method according to claim 1 or 3, wherein the device number is a device number at the time of fiber channel connection.

Further, the above object is achieved by a connection device recognition method according to any one of claims 1 to 5, wherein a plurality of servers and a disk storage device are connected via a network.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an example of a server system to which a connection device recognition method according to the present invention is applied.

The server system shown in FIG. 1 comprises a plurality of servers operating under program control, here a first server 110 and a second server 120, and a server 11
0, 120 and a shared disk storage device (shared disk) 100.

Here, the server 110 is booted (bo
ot) the registry (boot information registry) 111 storing the information, and the server 120
21 respectively. The shared disk 100 includes a logical disk 101 to which a logical unit number (LUN) A is allocated, and a logical unit number (LU).
N) A logical disk 102 to which B is allocated.

FIG. 2 is a flowchart showing the configuration of a device table creation process according to the present invention applied to the above-described server system. That is, as shown in FIG. 2, in the device table creation process of the present invention, a step 201 for reading the registries 111 and 121 storing the ID and LUN information of the device to be searched, here the SCSI device, is performed. Registry 111, 121
Step 2 of creating search table based on stored information of
02, a step 203 for setting a SCSI device number (SCSI ID) and LUN information according to the created search table, and a test unit ready to check whether a valid device is connected to the designated address (step 203). (Step 204) for issuing a TestUnit Ready command, a step 205 for confirming whether or not a selection timeout occurs when a device is not connected, and a step 206 for creating a device table for storing and managing information on connected devices.
And the next SCSI ID and LU according to the search table.
The method includes a step 207 for setting the N number and a step 208 for determining whether or not all the information set in the search table has been searched.

In the method of the present invention, the ID of the device to be searched, here the SCSI ID and LUN information, are stored in the registries 111 and 121 in advance, and the stored registry information is first read (step 20).
1) A search table is created based on the information (step 202).

Then, according to the created search table, the SCSI device number (SCSI ID) and the LU
N information is set in the system (step 203), and a test unit ready command is issued to confirm whether a valid device is connected to the specified address (step 204).

Next, if a valid device is not connected to the address to which the command is issued, a selection timeout occurs, so it is confirmed whether or not a selection timeout has occurred (step 205).

If no selection timeout has occurred (NO in step 205), that is, if there is a response from the device, a device table for storing and managing information on the connected device is created (step 206).

At step 205, when a selection timeout occurs (when step 205 is YES), and after the device table is created at step 206, the next SCSI ID is set according to the search table.
And LUN information are set in the system (step 20).
7).

Thereafter, it is determined whether or not all the information set in the search table has been searched (step 208). If there is still a valid device to be searched (step 208 is NO), step 203 is executed.
Thereafter, steps 203 to 207 are performed until all the information set in the search table is searched.
Is repeated.

As can be seen from the above-described device table creation process, in the connection device recognition method in the shared disk type server system of the present invention, each server 11
0, 120 registers the ID and LUN of the device to be accessed in the shared disk 100 in the boot information in advance, and searches for only the ID and the registered logical unit number and the logical unit (LU) at system startup. To make their system aware.

In the above embodiment, the device is SC
Although the description has been given of the case of the SI device, the present invention is not limited to this device alone. For example, the present invention can be applied to the recognition of a connected device in an environment using a device number such as a fiber channel connection. May be applied. Further, the plurality of servers 110, 120,... And the shared disk 100 may be connected via a network.

[0030]

As described above, according to the present invention, the following effects can be obtained. First, an effect of being able to search for a device with an arbitrary device number (including an LU here and the same hereinafter) without defining a device number (including a LUN here and the same hereinafter) from “0”. There is. Thereby, even if the number of the device to be acquired is particularly random for an LU in a cluster system or the like, an arbitrary device can be searched by reflecting it in the search table of the registry, and a device table can be created. In the conventional method, since the device number is searched in order from “0”, the subsequent device is searched when “0” does not exist, or when a device with a lower device number has a failure and cannot respond normally. A normal device having a number cannot be recognized, and the device cannot be detected. However, the present invention can avoid such a situation.

Second, since only necessary devices can be searched, high-speed device search can be performed. In the conventional method, a search for an unconnected device takes a long time to recognize whether or not the device is unconnected because of waiting for a selection timeout. However, in the present invention, only a necessary device can be searched by initial setting. This is because it does not take long to recognize whether or not.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example of a server system to which the method of the present invention is applied.

FIG. 2 is a flowchart for explaining an embodiment of the method of the present invention in the server system shown in FIG. 1;

FIG. 3 is a flowchart for explaining a conventional method.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 shared disk (shared disk storage device) 101, 102 logical disk 110 first server 111, 121 boot information registry 120 second server

Claims (6)

[Claims] 1. A server system having a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device being capable of defining a plurality of device numbers. A method of recognizing a connected device, wherein a number of a device to be targeted is registered in the boot information in advance, and at the time of system startup, only the registered device is searched and recognized by its own system. 2. A server system comprising a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device being capable of defining a plurality of logical unit numbers. The logical unit number to be accessed is registered in the boot information in advance, and when the system starts up,
A method for recognizing a connected device, wherein only a logical unit registered with a logical unit number is searched and recognized by its own system. 3. A server system comprising a plurality of servers and a disk storage device shared by the servers, each server having boot information, and the disk storage device being capable of defining a plurality of device numbers and a plurality of logical unit numbers. In, the number of the device and the logical unit number that each server wants to access is registered in the boot information in advance,
A connected device recognizing method characterized in that at the time of system start-up, only a device and a logical unit registered with a number and a logical unit number are searched and recognized by the own system. 4. The connected device recognition method according to claim 1, wherein the device is a SCSI device. 5. The connection device recognition method according to claim 1, wherein the device number is a device number when a fiber channel is connected. 6. The connection device recognition method according to claim 1, wherein the plurality of servers and the disk storage device are connected via a network.