JP2002244922A - Network storage system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a network storage system capable of storing data in a reliable storage device desired by a user. SOLUTION: In the network storage system controlling a plurality of storage devices connected with a network, reliability level information is given to data in the storage device, and the reliability level information is updated within a scope of access right of a client when the client refers to or updates the data in the storage device to change the storage device storing data based on the updated reliability level information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a computer system and a storage system, a network system connecting them, and a computer network system which is an integrated form thereof.

[0002]

2. Description of the Related Art As a storage system via a network, a SAN (Storage Area Network) or NAS
(Network Attached Storage), etc.
A model called Storage Virtualization, which builds a virtual storage system on a network, has also been announced.

[0003]

The purpose of the storage systems connected to these networks is to provide the data held by each storage system efficiently. Therefore, a storage system that holds data is predetermined, and the reliability of the data depends on the configuration of each storage system.
As a method of guaranteeing data reliability, there is a construction of RAID. RAID refers to David A. Patterson, Garth Gibson, and Randy Etch
Kutz co-authored "Inexpensive Disk Redundancy Array (Redund
antArrays of Inexpensive Disks, RAID) "University of California Report U
It is described in CB / CSD87 / 391 (December, 1987).
It is classified into six types up to AID5.

On the other hand, a user who accesses data via a network can determine how reliable individual data is stored in a storage system.
I cannot know directly. If necessary data is stored in a single hard disk in the server, if the server goes down, the data cannot be used. If the data is on a single disk, there is a risk of data loss. Therefore, the user must back up the necessary data on the network by himself / herself in case of an unexpected situation, and backup frequently updated data at the same cycle as the update cycle. Must be done.

[0005] Repeating these backup operations wastes the bandwidth of the network. Also, as a conventional technique, there is a technique called caching, in which frequently used data is copied to a network server or storage system that is physically close to a user or a client, which is easily accessible. , And does not necessarily match the need for data. further,
Although rarely referenced, caching does not work for data retention for highly important data.

An object of the present invention is to provide a network storage system capable of storing data in a storage of a reliability level desired by a user.

[0007]

In order to solve the above-mentioned problems, the present invention provides a network storage system having a plurality of storage devices connected to a network. And when the client connected to the network refers to or updates the data on the storage device, updates the reliability level information within the range of the client's access authority, based on the updated reliability level information. The storage device for storing data can be changed. In this way, data can be stored in the storage device of the reliability level desired by the user (client).

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the illustrated embodiment. (1) Outline Description of the Present Invention First, the overall operation of the present invention will be described based on the flowchart shown in FIG. As shown in FIG. 1, in a network storage system according to the present invention (see FIG. 2), when a client refers to or updates data on the storage system, the client does not trust the data, the storage system, or the network server. A request for reliability or the like is notified to the storage system on the network as reliability level information (step S1).

The network storage system determines whether the access permission level of the client is appropriate (step S2). If the access permission level is not appropriate, the network storage system rejects the request from the client. Notify that there is no (step S
3). If the access permission level of the client is appropriate, the storage system on the network changes the reliability level information of the data to a level requested by the client according to the access permission level set for each client ( Step S4).

The storage system that has received the notification changes the reliability level information of the data and, at the same time, checks whether a data storage form that matches the reliability level information is possible in its own storage system (step S).
5). If a storage mode that matches the reliability level information is available in its own storage system, the storage mode of the data is changed, and the process ends (step S6).
If a storage mode that matches the reliability level information is not possible in its own storage system, a storage system that uses a storage mode that matches the reliability level information is searched for using a network search engine (step S7).

According to the search result, if there is a storage system available on the network that can use the storage mode matching the reliability level information, the data is copied to the storage system (step S8). After receiving the copy completion notification from the copy destination storage system, the search source storage system deletes the data in its own storage system and notifies the entire network storage system of the data movement (step S9). The client that has requested the change of the reliability level information is notified via the network storage system that the storage system for storing the data has been changed (step S10).

According to the search result, if there is no storage system on the network that can use the storage mode that matches the reliability level information, the search source storage system updates the reliability level information via the network storage system. In addition to notifying that it is impossible, the data reliability level information is restored (step S11).

(2) Embodiment FIG. 1 is a system configuration diagram of this embodiment, which is a general network storage system. In FIG. 1, clients (computers, workstations) C1, C2, servers (servers, minicomputers, mainframes) S1, S2, S3, and disk arrays A2, A3, A4 are directly connected on a network N1. ing.
The disk array A1 is connected under the server S1, the disk arrays 51 to 54 are connected under the server S3, and mirroring is performed by the processing of the server 3, so that the disk array A5 is equivalently installed on the network.
Are considered to be connected.

FIGS. 3 to 7 show the operation of this embodiment. In FIG. 3, the client C1 sends the disk array A
4, a reference request for data D1 is issued. At this time, the client C1 changes the reliability level information of the data D1 to “RA
A request is issued to change from “ID0” to “RAID1.” In FIG. 4, the disk array A4 itself requests the client C1 “change from RAID0 to RAID1”.
Therefore, a copy request of the data D1 is made to the disk array A2 using the search engine.
In this embodiment, the disk array A2 rejects the copy request due to a problem such as a free disk space.

In FIG. 5, the disk array A4 is connected to the server S3.
Request a copy of the data D1. Server S3
Is connected to a mirrored disk array under its control, and is regarded as equivalent to a RAID1 disk array as a network storage system.
In response to the copy request from the disk array A4, the server S3 receives the data D1 of the disk array A4 to copy the data D1 to the subordinate mirrored disk array A5.

In FIG. 6, the server S3 has a disk array A4
The data D1 received from the disk array A
A write instruction is issued to the A 51 and the A 52. The server S3 that has received the write completion report from the disk arrays A51 and A52 notifies the disk array A4 of the write completion (copy completion).

In FIG. 7, the disk array A4 is connected to the server S3.
Receiving the write completion notification from the server, notifies the other network storage systems on the network of the movement of the data D1 and deletes the data D1. Also,
The disk array A4 notifies the client C1 that the update of the reliability level information has been completed.

FIG. 8 shows an example of a case where the change of the reliability level information is rejected. When referring to the data D2 on the disk array A4, the client C2 requests to change the reliability level information to "RAID5". The disk array A4 compares the user permission level of the data D2 with the user level of the client C2. In this embodiment, the user level of the client C2 is "LEVEL".
2 ", whereas the user permission level of the data D2 is" LEVEL4 ", it is determined that the user level is lower, and a rejection of the reliability level information update request from the client C2 is rejected. Make a notification.

FIG. 9 shows a data management method on each storage system. In FIG. 9, data D
1 to D4, reliability level information R1 to R4
Is associated with. The reliability level information holds a RAID level that defines the reliability and a user permission level that determines whether the user can change the reliability level.

In this embodiment, the higher the user permission level and the higher the user level, the higher the priority. In this embodiment, the lower the priority, the higher the priority. Further, in the present embodiment, the RAID level of the disk array is exemplified as the reliability-related information indicated by the reliability level information. However, in addition to the RAID level, the ID information of the hard disk vendor, the physical capacity of the hard disk, and the defectiveness The amount of the truck, the failure history, and the like can be added to the reliability level information as reliability information.

[0021]

As described above, according to the present invention, the following effects can be obtained. That is, the user can request that necessary data be stored in a storage system that guarantees necessary reliability, so that unnecessary backup is not required. In addition, by presenting a request for data reliability from the user side, there is also an effect of saving the trouble of system management.

[Brief description of the drawings]

FIG. 1 is an overall flowchart of the present invention.

FIG. 2 is a system configuration diagram of a network storage system according to an embodiment of the present invention.

FIG. 3 is “Part 1” of the operation of the network storage system, and is an operation of requesting an update of reliability level information when a client refers to data.

FIG. 4 is the “Part 2” of the operation, and is the disk array A
The operation No. 4 requests the disk array A2 to copy the data D1 and is rejected.

FIG. 5 is “Part 3” of the operation, which is the disk array A
4 requests the disk array A5 to copy the data D1 and is accepted. The operation of the server S3 is equivalently regarded as the disk array device A5 on the network storage system.

FIG. 6 is “Part 4” of the operation, which is the disk array A
This is an operation of notifying the disk array A4 of a write completion report of the data D1 to the disk array A4.

FIG. 7 is a “No. 5” of the operation, in which the disk array A
4 is an operation for deleting the data D1 and notifying that the data has moved to the network.

FIG. 8 is “No. 6” of the operation, and the client C2
Therefore, the request for updating the reliability level information for the data D2 is rejected.

FIG. 9 is a description example of reliability level information in the embodiment of the present invention.

[Explanation of symbols]

 N1 Network A1-A5, A51-A54 Disk Array C1 Computer C2 Workstation S1 Server S2 Minicomputer S3 Mainframe

Claims (6)

[Claims] In a network storage system having a plurality of storage devices connected to a network, reliability level information is provided to data on the storage device, and a client connected to the network is provided with a client on the storage device. When referencing or updating data, it is possible to update the reliability level information within the range of the client's access authority, and to change the storage device that stores the data based on the updated reliability level information. A network storage system, characterized in that: 2. The network storage system according to claim 1, wherein said storage device is directly connected to said network. 3. The network storage system according to claim 1, wherein said storage device is connected to a network server system directly connected to said network. 4. The storage device according to claim 1, wherein the network device system directly connected to the network and the storage device connected to the network server system are regarded as a single storage system. 2. The network storage system according to 1. 5. The network storage system according to claim 1, wherein the storage device includes one or more disk array devices. 6. The storage device according to claim 4, wherein the storage device is a network server system connected to one or a plurality of storage devices, and includes a storage system considered as a disk array system as a whole. Network storage system.