Method for storing data using globally distributed storage system, and program and storage medium for allowing computer to realize the method, and control apparatus in globally distributed storage system

• US 20040064633A1
• Filed: 09/22/2003
• Published: 04/01/2004
• Est. Priority Date: 09/30/2002
• Status: Abandoned Application

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Oota, Masao
• Application Number:
  US10/665,133
• Publication Number:
  US 20040064633A1
• Time in Patent Office: Days
• US Class Current: 711/100
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0631 by allocating resources to storage systems
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A control device for preparing redundant data, dividing the data into a plurality of volumes, and distributing and storing volumes in a plurality of storage units scattered through a network includes a route management unit and a storage set management unit. The route management unit computes an evaluation value indicating the preferability of a use target on each of the scattered storage unit based on the bandwidth, the communications cost, the physical distance between a node requesting a write and a storage unit. The storage set management unit selects a plurality of storage units as the optimum storage set from among the above mentioned scattered storage units based on the evaluation value.

Claims

1. A method for storing data by a computer dividing redundant data into a plurality of volumes, distributing and storing the volumes in a plurality of scattered storage units through a network, comprising:

computing an evaluation value indicating desirability of each of the scattered storage units to be used based on a bandwidth, a communications cost, and a physical distance between a node to which a write request is issued and a storage unit; and

selecting a plurality of storage units as an optimum storage set from among the scattered storage units based on the evaluation value.

2. The method according to claim 1, wherein

in computing the evaluation value, a hop count from a node to which the write request is issued to each storage unit is further used.

3. The method according to claim 1, further comprising

providing the storage unit as a virtual storage unit for a user of the system.

4. The method according to claim 1, further comprising:

when the data is read from the storage set, reading from each storage unit volumes not containing a redundant portion among the plurality of volumes written to the storage set; and

reconstituting the data using the read volumes.

5. The method according to claim 3, further comprising:

when the data is read, computing a use priority indicating high response based on the bandwidth and the cost for each storage unit; and

determining which volumes among the plurality of volumes are to be read from each storage unit as volumes not containing a redundant portion based on the use priority.

6. The method according to claim 1, further comprising

storing a replica of a first volume in the plurality of volumes in a storage unit not selected as the storage set.

7. The method according to claim 6, further comprising

when a replica of the first volume is generated, based on the evaluation value selecting a method from between two generating method based on the evaluation value, that is, replicating the first volume from a storage unit storing the first volume, and regenerating the first volume using redundancy from volumes other than the first volume in the plurality of volumes.

8. The method according to claim 6, further comprising

selecting a storage unit storing a replica of the volume from among the storage units not selected as the storage set based on the evaluation value.

9. The method according to claim 6, further comprising

writing a volume in a multicast system to a plurality of storage units storing the same volume.

10. The method according to claim 6, wherein

when a replica of the first volume is written to a storage unit, a writing process is performed in plural operations.

11. The method according to claim 1, further comprising

when a fault occurs in the first storage unit in the storage set, limiting a write to other storage units in the storage set.

12. The method according to claim 1, further comprising

when a fault occurs in a third storage unit in the storage set, selecting based on the evaluation value a fourth storage unit other than a storage unit selected as the storage set instead of the third storage unit.

13. The method according to claim 1, further comprising:

after selecting the storage set, reselecting a storage set in each node at a predetermined timing; and

when there is a volume not used by any node as a result of reselection, deleting the volume from a storage unit.

14. The method according to claim 13, wherein

the predetermined timing refers to a predetermined period after previous selection or a timing of changing a state of a volume.

15. The method according to claim 1, further comprising:

after reading the data, temporarily storing the data for a predetermined period in an arbitrary storage unit; and

when data is read within the predetermined period, reading temporarily stored data from the storage unit.

16. A method according to claim 1, further comprising:

temporarily storing data specified in a write request within a predetermined period in a temporary storage area;

retrieving data from the temporary storage area after the predetermined period;

dividing the data into a plurality of volumes; and

writing the plurality of volumes in the storage set.

17. The method according to claim 15, further comprising

when a reading or writing process is performed on data including the temporarily stored data, reading or writing only a portion of data not containing the temporarily stored data.

18. The method according to claim 1, further comprising

when the plurality of volumes are written to the storage set, prohibiting a writing process on the storage set until a write is completed for a node to which the write request is issued.

19. The method according to claim 18, further comprising

determining a storage unit as a representative storage unit from among a plurality of storage units storing the same volumes, wherein:

in prohibiting a writing process in the plurality of storage units,

prohibiting a writing process to the representative storage unit is performed by a node to which the write request is issued; and

prohibiting a writing process to a storage unit other than the representative storage unit is performed by the representative storage unit.

20. The method according to claim 19, wherein

the representative storage unit is a storage unit for storing a volume as original data.

21. A computer-readable storage medium storing a program used to direct a computer to control processes of dividing redundant data into a plurality of volumes in a system comprising storage units scattered through a network, and distributing and storing volumes in a plurality of storage units, comprising:

computing an evaluation value indicating desirability of each of the scattered storage units to be used based on a bandwidth, a communications cost, and a physical distance between a node to which a write request is issued and a storage unit; and

selecting a plurality of storage units as an optimum storage set from among the scattered storage units based on the evaluation value.

22. A control device which controls dividing redundant data into a plurality of volumes in a system having scattered storage units through a network, and distributing and storing volumes in a plurality of storage units, comprising:

route management means for computing an evaluation value indicating desirability of each of the scattered storage units to be used based on a bandwidth, a communications cost, and a physical distance between a node to which a write request is issued and a storage unit; and

storage set management means for selecting a plurality of storage units as an optimum storage set from among the scattered storage units based on the evaluation value.

23. A control device which controls dividing redundant data into a plurality of volumes in a system having scattered storage units through a network, and distributing and storing volumes in a plurality of storage units, comprising:

a route management unit computing an evaluation value indicating desirability of each of the scattered storage units to be used based on a bandwidth, a communications cost, and a physical distance between a node to which a write request is issued and a storage unit; and

a storage set management unit selecting a plurality of storage units as an optimum storage set from among the scattered storage units based on the evaluation value.