JP3534302B2 - Multipath system for storage devices - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a path for accessing a specific storage device (access path between a host computer and a storage device. For example, a combination of a driver, a bus control adapter, a control bus, and a device controller. Access path)
The present invention relates to a multi-path system of a storage device which is a computer system having a plurality of types.

[0002]

2. Description of the Related Art In a storage device multipath system which is a computer system for controlling storage devices, each storage device must be uniquely identified in order to specify an access path to the storage device.

Conventionally, in a multipath system of this type of storage device, each access path is detected by using a unique ID (IDentification) given to each storage device in advance in the computer system.

For example, in a computer system in which the fiber channel protocol is used, each device has a WWN (World Wide Name), which is a unique ID in the world.
Each storage device can be uniquely identified by using it.

In addition, SCSI (Small Compu
Even in computer systems where the ter Systems Interface) protocol is used, by supporting vendor-unique commands,
A unique ID for the computer system can be given to each device, and each storage device can be uniquely recognized.

[0006]

However, in order to detect each access path by the above conventional technique, it is a condition that each storage device always has a unique ID. On the other hand, general storage devices, particularly SCSI disk devices, do not have a unique ID in most cases.

Therefore, in the conventional multipath system for storage devices, it is difficult to properly detect (recognize) a plurality of access paths for one storage device, and the access load for one storage device becomes a plurality of access paths. There was a problem that it was difficult to disperse properly.

In the present invention, by giving a unique ID for identifying a storage device from the host computer side, it is possible to uniquely recognize the storage device, and a plurality of access paths for each storage device are provided. To provide a multipath system of a storage device capable of realizing the detection of the above.

That is, the object of the present invention is to enable automatic detection of a plurality of access paths to one storage device and to appropriately distribute the load of access to the storage device to each access path. To provide a multipath system of a storage device having a configuration.

Japanese Patent Publication No. 6-54481 is a patent publication relating to the prior art for the present invention.

The technique (data processing system) described in this publication has path information creating means in the CPU, and the path information creating means creates access path information connecting the channel device and the magnetic disk device.

However, this conventional technique has a problem that it cannot be applied to a computer system in which a general magnetic disk device is controlled by a host computer.

The reason why such a problem exists is that a computer system based on the peculiarity that this technology "preliminarily recognizes the positional relationship between the channel device and the magnetic disk device as hardware information" is assumed. This is because

[0014]

A multipath system for storage devices according to the present invention is a unique storage area in each storage device, which is an area for storing a unique ID given to each storage device by a host computer. Using the ID storage area and the access path between the host computer and the storage device, a plurality of paths in the host computer for reading data of a predetermined size from the unique ID storage area of the storage area of the storage device are read. The unique ID storage area data reading unit in the means and the format of the data read from the storage device by the unique ID storage area data reading unit are unique I
New unique ID if different from D format
And the new unique ID is read by the unique ID writing unit in the multiple path detecting means in the host computer that writes the new unique ID in the unique ID storage area in the storage device, and the unique ID storage area data reading unit. Unique data format is unique
If the format is, the unique ID and the current search path are mapped and the unique I
When the format of the data read by the D storage area data reading unit is different from the format of the unique ID, the unique ID newly assigned by the unique ID writing unit is mapped to the search path, And a mapping unit in the multipath detecting means in the host computer that holds mapping data having information in the host computer.

[0015]

Further, such a multi-path system for storage devices is such that each storage device has a unique ID storage area, which is an area for storing a unique ID given to each storage device by the host computer, in the storage area. the host computer connected via a plurality of access paths, Hosutoko
Access path between the computer and the storage device
Using the storage device storage area
A predetermined size of data from the
Unique ID storage area data reading unit for reading data,
Read by the unique ID storage area data reading unit
The format of the included data is a unique ID former
If a new unique ID is assigned,
The new unique ID is stored in the storage device
Write unique ID in unique ID storage area
Section and the unique ID storage area data reading section
More data read from the storage device
If the format is a unique ID format
Mapping the unique ID to the current search path
And the unique ID storage area data reading unit
The format of the read data is the unique ID
If the format is different, write the unique ID
The unique ID newly assigned by the department and the search
Performs mapping with the path and stores the information of the mapping.
Mapping data to be stored in the host computer
It is also possible to realize it as a recording medium recording a program for functioning as a mapping unit .

Further, the storage device multi-path system according to the present invention refers to the mapping data generated / held by the control of the multi-path detecting means to balance the load of a plurality of access paths to each storage device. It is also possible to add the access path load balancing means.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, the present invention will be described in detail with reference to the drawings.

(1) First Embodiment FIG. 1 is a block diagram showing the configuration of a multipath system for storage devices according to the first embodiment of the present invention.

Referring to FIG. 1, a multipath system for storage devices according to the present embodiment includes a host computer 1 operating under program control, and bus control adapters 2 and 3 connected to the host computer 1.
, Control buses 4 and 5, and a storage device 6. Although not shown in FIG. 1, a plurality of storage devices including the storage device 6 exist in the multipath system of the storage device.

The storage device 6 comprises device controllers 61 and 62 and a storage area 600 having a unique ID storage area 60.

The unique ID storage area 60 is a storage area 6
A unique ID (ID for identifying the storage device 6) having a predetermined size and having a predetermined format is stored at a predetermined address location in 00 (from the beginning) The unique ID is not stored, but the unique ID is given by the control of the host computer 1).

The host computer 1 includes a driver 110 for controlling the bus control adapter 2 and a bus control adapter 3
And a plurality of access paths (for example, the access paths 400 and 401 for the storage device 6) to each storage device (for example, the storage device 6) are detected, and these are the access paths for the storage device. And a plurality of path detecting means 120 that associate (map) with each other. It should be noted that the access path load balancing means 13 shown by a broken line box in the host computer 1
Zero will be described later.

The multiple path detecting means 120 uses a unique ID.
The storage area data reading unit 121, the unique ID writing unit 122, and the mapping unit 123 are included.

The respective units in the multi-path detecting means 120 generally operate as follows.

Unique ID storage area data reading unit 12
1 uses the access path between the host computer 1 and the storage device (for example, the access paths 400 and 401 between the host computer 1 and the storage device 6) to store the unique ID storage area of the storage area of the storage device. (For example, the storage device 6
The data of a predetermined size is read from the unique ID storage area 60) of the storage area 600.

The unique ID writing unit 122 is newly added when the format of the data read by the unique ID storage area data reading unit 121 is different from the predetermined format, that is, the format of the unique ID determined by the host computer 1. A unique ID and assigns the new unique ID to the unique ID storage area (the unique ID in which the data is read by the unique ID storage area data reading unit 121).
D storage area. For example, it writes in the unique ID storage area 60). The unique ID storage area data reading unit 1
If the format of the data read by 21 is the unique ID format, no new unique ID is written.

When the format of the data read by the unique ID storage area data reading section 121 is the unique ID format, the mapping section 123 determines the unique ID and the current search path (the access used in the above access). Path, that is, an access path to be detected), and if the format of the data read by the unique ID storage area data reading unit 121 is different from the unique ID format, a new number is assigned by the unique ID writing unit 122. The generated unique ID is mapped to the search path, and mapping data having information on the mapping is held in the host computer 1. By this mapping, for example, an access path 400 and an access path 401 having a common unique ID (unique ID for identifying the storage device 6) to be mapped are set as a plurality of access paths for one storage device 6. Will be recognized by the host computer 1.

FIG. 2 is a flow chart showing the processing (processing at the time of access path detection / mapping data generation) of the multi-path system of the storage device according to this embodiment. This process is the first access path search step A1.
And the unique ID storage area data reading step A2
Format determination step A3, unique ID new numbering step A4, unique ID writing step A
5 and unique ID / search path mapping step A
6, a next access path search step A7, and a search path presence / absence determination step A8.

Next, the operation of the multipath system for the storage device according to the present embodiment configured as described above will be described in detail.

The multiple-path detecting means 120 in the host computer 1 detects a plurality of access paths, and each storage device (for example, the storage device 6) and a plurality of access paths (for example, a storage device) for accessing the storage device. In order to associate (map) the access paths 400 and 401) for accessing the device 6, the following processing is performed (see FIG. 2).

First, the unique ID storage area data reading unit 121 in the multipath detecting means 120 searches (detects) the first access path. That is, unique I
The D storage area data reading unit 121 uses one access path (here, the access path 400 passing through the driver 110, the bus control adapter 2, the control bus 4, and the device controller 61 in FIG. 1) as the first access path.
Is detected) (step A1 in FIG. 2). In addition, the access path detected in this way is called a "search path".
Call.

Next, the unique ID storage area data reading unit 121 uses the access path 40 detected in step A1.
By using 0 (search path), a specific address area (unique ID in the storage area 600 of the storage device 6)
The data (specified data) stored in the storage area 60) is read (step A2).

Unique ID in the multiple path detection means 120
The writing unit 122 determines (inspects) whether the format of the data read by the unique ID storage area data reading unit 121 is the unique ID format (correct format) (step A3). That is, it is determined whether or not the data is the data written by the multi-pass detection unit 120 (that is, the unique ID).

In this case (when the access path 400 is a search path), the data in the unique ID storage area 60 is not a unique ID, so the determination result in step A3 is "NO".

When the unique ID writing unit 122 determines that "the data is not in the correct format (unique ID format)" in step A3 as described above, it newly assigns a unique ID (step A
4) Write the new unique ID in the unique ID storage area 60 of the storage area 600 of the storage device 6 (step A5).

The mapping section 123 in the multi-path detecting means 120, after finishing the processing of step A5, executes step A
Unique ID written in 5 (storage device 6
, And the search path at this time (here, the access path 400) are mapped,
The mapping data having the mapping information is stored (held) in the host computer 1 (step A).
6).

Then, the unique ID storage area data reading unit 121 in the plural path detecting means 120 searches (detects) the next access path. That is, unique I
The D storage area data reading unit 121 attempts to detect one unprocessed access path as the next access path (step A7), and whether or not the access path can be detected in step A7 (whether or not there is a search path). ) Is determined (checked) (step A8). Here, it is assumed that the access path 401 passing through the driver 111, the bus control adapter 3, the control bus 5, and the device controller 62 in FIG. 1 is detected as a search path.

Unique ID storage area data reading unit 12
When it is determined in step A8 that “there is a search path” (corresponding to this case in this case) by step 1, the multi-path detecting means 120 performs step A2 similarly to the above.
The following processing is executed.

That is, the unique ID storage area data reading unit 121 uses the access path 401 (search path) detected in step A7 to store the storage device 6
The data stored in the unique ID storage area 60 in the storage area 600 is read (step A2).

Unique ID in the multiple path detection means 120
The writing unit 122 determines whether the format of the data read by the unique ID storage area data reading unit 121 is the unique ID format (step A3).

In this case (when the access path 401 is the search path), since the data in the unique ID storage area 60 is the unique ID, the determination result of step A3 is "YES". That is, the access path 401
Unique ID storage area 6 whose data was read via
Since the unique ID has already been written in 0 through the access path 400 as described above, the format of the data to be determined is the unique ID format.

When the unique ID writing unit 122 determines in step A3 that "the data has the correct format (unique ID format)" as described above, it does not write the unique ID, Control is passed to the mapping unit 123 in the detection means 120.

The mapping unit 123 reads the data (unique ID corresponding to the storage device 6) read in step A2 and the current search path (here,
The access path 401) is mapped and the mapping data having the mapping information is stored (held) in the host computer 1 (step A6).

When it is determined in step A8 that "there is no search path", the plural path detecting means 120 is used.
Ends the process shown in FIG.

In this way, the host computer 1
Is the mapping data generated by the multiple path detecting means 120 (for example, access paths 400 and 401).
And the storage device 6 can hold mapping data having information associated with each other. As a result, the host computer 1 recognizes that the access path 400 and the access path 401 are a plurality of access paths to the same storage device 6, and based on the recognition, the load of access to the storage device 6 is determined. The access paths 400 and 401 can be appropriately distributed.

Specifically, as shown by the broken line box in FIG. 1, the mapping data generated and held under the control of the multipath detecting means 120 is referred to, and the storage devices (for example, the storage device 6) are referenced. Access paths (eg, access paths 400 and 40)
Access path load balancing means 130 for performing load balancing in 1)
May be a component of the host computer 1.

(2) Second Embodiment FIG. 3 is a block diagram showing the configuration of a multipath system for storage devices according to the second embodiment of the present invention.

Referring to FIG. 3, the storage device multipath system according to the present embodiment comprises a host computer 1001 and an FC-AL (Fiber Cha).
nnel-Arbitrated Loop) control unit 1
301 to 1309, FC-AL (Fiber Channel Local Loop) 1401 to 1403, and a storage device 2001. Note that FIG.
Although not shown in the figure, there are a plurality of storage devices including the storage device 2001 in the multipath system of the storage device.

The host computer 1001 comprises a plurality of path detecting means 1100 and drivers 1201 to 1209 (access path load balancing means 13).
00 will be mentioned later).

The multipath detecting means 1100 has a unique I
The D storage area data reading unit 1101, the unique ID writing unit 1102, and the mapping unit 1103 are included.

The storage device 2001 includes device controllers 2101 to 2103 and a storage area 2100 having a unique ID storage area 2110.

The unique ID storage area 2110 is a unique ID (identifying the storage device 2001) which is data of a predetermined size and a predetermined format, at a predetermined address location in the storage area 2100. (The unique ID is not stored from the beginning, but the unique ID is given by the control of the host computer 1001).

In this embodiment, the host computer 10
The number of paths (access paths) from 01 to access the storage device 2001 is multiplexed, and the access paths are the driver, the FC-AL control unit, and the FC-A.
A combination of L and a device controller (for example, driver 1201, FC-AL control unit 1301, F
C-AL1401, and device controller 210
This is different from the first embodiment shown in FIG. 1 in that it is configured by a path passing through 1).

As described above, according to the present invention, the access path for accessing a specific storage device is shown in FIG.
It is possible to carry out the present invention not only in two ways as shown in (3) but also in N (any positive integer of 3 or more) extensions. Further, the access path is not limited to the mode (mode through the driver, bus control adapter, control bus, and device controller) as shown in FIG. 1, and the access path as shown in FIG. The present invention can be implemented by using an access path connected in a mesh pattern.

The multiple path detecting means 1100 (unique ID storage area data reading unit 1 in this embodiment
The operation contents of 101, the unique ID writing unit 1102, and the mapping unit 1103 are the same as those of the multiple path detecting unit 120 (the unique ID storage area data reading unit 121, the unique ID writing unit 122, in the first embodiment).
And the operation of the mapping unit 123) is the same.

Also in the present embodiment, it is possible to provide the same means as the access path load distribution means 130 in FIG. For example, as indicated by the dashed box in FIG. 3, each storage device (for example, the storage device 2001) is referenced by referring to the mapping data generated and held under the control of the multiple path detection unit 1100.
An access path load balancing means 1300 for performing load balancing of access paths to the access paths (for example, nine access paths between the host computer 1001 and the storage device 2001 shown in FIG. 3).
It is conceivable to make it one constituent element.

(3) Third Embodiment FIG. 4 is a block diagram showing the configuration of a multipath system for storage devices according to a third embodiment of the present invention.

Referring to FIG. 4, the multipath system for storage devices according to the third embodiment of the present invention is different from the multipath system for storage devices according to the first embodiment shown in FIG. The difference is that a recording medium 40 in which a multi-pass detection processing program is recorded is provided. The recording medium 40 may be a magnetic disk, a semiconductor memory, or another recording medium.

The multiple path detection processing program is connected from the recording medium 40 to the host computer 1 (each storage device (including the storage device 6)) via a plurality of access paths (including the access paths 400 and 401). Host computer), and the operation of the host computer 1 is detected by the multi-path detecting means 120 (unique ID storage area data reading unit 12).
1, the unique ID writing unit 122, and the mapping unit 123). The operation of the multi-pass detection means 120 under the control of the multi-pass detection processing program is performed by the multi-pass detection means 120 (unique ID storage area data reading unit 121, unique ID writing unit 122, and mapping unit 123) in the first embodiment. Since it is exactly the same as the operation of, the detailed explanation is omitted.

(4) Fourth Embodiment FIG. 5 is a block diagram showing the configuration of a multipath system for storage devices according to the fourth embodiment of the present invention.

Referring to FIG. 5, the storage device multipath system according to the fourth embodiment of the present invention is different from the storage device multipath system according to the second embodiment shown in FIG. The difference is that a recording medium 50 in which a multi-pass detection processing program is recorded is provided. The recording medium 50 may be a magnetic disk, a semiconductor memory, or another recording medium.

The multi-path detection processing program is read from the recording medium 50 into the host computer 1001 (host computer connected to each storage device (including the storage device 2001) via a plurality of access paths), and the host computer 1001. Computer 1001
The operation of the multi-path detecting means 1100 (unique ID storage area data reading unit 1101, unique ID writing unit 1
102, and the mapping unit 1103). The operation of the multi-pass detection means 1100 under the control of the multi-pass detection processing program is performed by the multi-pass detection means 1100 (unique ID storage area data reading unit 1101, unique ID writing unit 1102 in the second embodiment.
Since the operation is the same as that of the mapping section 1103), detailed description thereof will be omitted.

[0064]

As described above, according to the present invention,
The following effects are produced.

The first effect is that the same access path as the conventional one (for example, an access path consisting of a driver, a bus control adapter, a control bus, and a device controller) is used, and one operator's intervention is not required. It is possible to detect a plurality of access paths to a storage device in association with the storage device.

The reason why such an effect occurs is that a storage area of the storage device is provided with a unique ID storage area for storing the unique ID given by the host computer, and the access path to the storage device is provided by the unique ID. This is because the storage device is identified in association with the storage device.

The second effect is to access the storage device by using the information (mapping data) about the plurality of access paths detected by mapping (associating) with the storage device as described above. That is, it is possible to appropriately distribute the access load to a plurality of access paths so that only one access path to the storage device is not loaded.

Here, in order to realize such a second effect, for example, it is conceivable to provide access path load balancing means.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a multipath system for storage devices according to a first embodiment of the present invention.

FIG. 2 is a flow chart showing processing (processing at the time of access path detection / mapping data generation) of the multipath system of the storage device shown in FIG.

FIG. 3 is a block diagram showing a configuration of a multipath system for storage devices according to a second embodiment of the present invention.

FIG. 4 is a block diagram showing the configuration of a multipath system for storage devices according to a third embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of a multipath system for storage devices according to a fourth embodiment of the present invention.

[Explanation of symbols]

1, 1101 Host computer 2, 3 Bus control adapter 4, 5 Control bus 6, 2001 Storage device 40, 50 Recording medium 60, 2110 Unique ID storage area 61, 62, 2101-2103 Device controller 110, 111, 1201-1209 Driver 120, 1100 Multiple path detecting means 121, 1101 Unique ID storage area Data reading section 122, 1102 Unique ID writing section 123, 1103 Mapping section 130, 1300 Access path load balancing section 400, 401 Access path 600, 2100 Storage area 1301-1309 FC-AL control units 1401-1403 FC-AL

Front page continued (58) Fields surveyed (Int.Cl. ⁷ , DB name) G06F 13/10-13/14 G06F 3/06

Claims (5)

(57) [Claims] 1. An access between a host computer and a storage device, and a unique ID storage area in a storage area in each storage device which is an area for storing a unique ID given to each storage device by the host computer. A unique ID storage area data reading unit in a plurality of path detecting means in the host computer for reading data of a predetermined size from the unique ID storage area of the storage area of the storage device using the path; When the format of the data read from the storage device by the storage area data reading unit is different from the format of the unique ID, a new unique ID is assigned and the new unique ID is stored in the unique ID in the storage device. If the format of the data read by the unique ID writing unit in the multiple path detecting means in the host computer writing in the area and the unique ID storage area data reading unit is the unique ID format, the unique ID and the current If the format of the data read by the unique ID storage area data reading unit is different from the format of the unique ID by performing mapping with the search path, the unique ID newly assigned by the unique ID writing unit and the search A multipath system for a storage device, comprising: a mapping unit that performs mapping with a path and holds mapping data having information of the mapping in a host computer in a plurality of paths detecting means. . 2. The generation by control of the multipath detecting means
Multipath system of the storage device according to claim 1, characterized in that it has an access path load balancing unit in the host computer refers to the mapping data stored for load balancing of the plurality of access paths for each storage device. Wherein access path drivers, bus control adapter, control bus, and claim 1 or a multi-pass system of the storage device according to claim 2, characterized in that a path through the device controller. 4. The storage device has a unique ID storage area, which is an area for storing a unique ID given to each storage device by a host computer, and is connected to each storage device via a plurality of access paths. The host computer uses the access path between the host computer and the storage device to store the unique I of the storage area of the storage device.
A unique ID storage area data reading unit that reads data of a predetermined size from the D storage area, and a new unique when the format of the data read by the unique ID storage area data reading unit is different from the unique ID format The format of the data read from the storage device by the unique ID writing unit that assigns an ID and writes the new unique ID to the unique ID storage area in the storage device and the unique ID storage area data reading unit is unique. When the ID format is used, the unique ID is mapped to the current search path, and the format of the data read by the unique ID storage area data reading unit is different from the unique ID format. Is a program for performing mapping between the unique ID newly assigned by the unique ID writing unit and the search path, and causing the mapping data having the information of the mapping to function as a mapping unit that is held in the host computer. Recording medium recorded. 5. The recording medium according to claim 4 , wherein the access path is applied to a multipath system of a storage device, which is a path passing through a driver, a bus control adapter, a control bus, and a device controller.