Multiple processor data processing system with mirrored data for distributed access

• US 6,542,962 B2
• Filed: 03/21/2001
• Issued: 04/01/2003
• Est. Priority Date: 03/21/2001
• Status: Expired due to Fees

• Current Assignee: Hitachi America Limited (Hitachi, Ltd.), Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Yamamoto, Akira, Kodama, Shoji
• Primary Examiner(s): Nguyen, Hiep T.
• Application Number:
  US09/815,494
• Publication Number:
  US 20020138696A1
• Time in Patent Office: 741 Days
• Field of Search: 711/114, 711/148, 711/153, 711/173, 709/214, 709/215, 714/6
• US Class Current: 711/114
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/10015 Access to distributed or replicated servers, e.g. using brokers
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

A data storage includes multiple disk units accessible to multiple processors/servers. The multiple disk units include a master disk unit and a one or more data-mirroring disk units. A data-mirroring disk unit is assigned to a corresponding ones of the multiple servers by one of the processors designated as the mount manager. Data is written by the processors to the data storage is written to the master disk unit, and copied by the data storage to the data-mirroring disk units. Data is read by each of the processors from the data-mirroring disk unit assigned to such processor.

Claims

1. A processing system having a plurality of disk units communicatively connected to two or more server processors by a storage system a method of distributing read access to data stored on the plurality of disk units that includes the steps of:

identifying one of the plurality of disk units as a master disk unit

assigning each of the other of the plurality of disk units to a corresponding one of the two or more server processors;

writing data received from the two or more server processors to the master disk unit;

copying the data to the other of the plurality of disk units;

receiving at the storage unit a request to read data from one of the other processor units to read data from the one of the other of the plurality of disk units assigned to the one processor unit and send the data to the one processor unit.

2. The method of claim 1, including the steps of writing data to a first location of the master disk unit;

and

before the first data is copied to a one of the other of the plurality of disk units, receiving a request to read data from a location of the one disk unit corresponding to the first location; and

reading the data from the first location of the master disk unit and sending the data to the server processor.

3. The method of claim 1, including the step of maintaining at each of the server processors a mount point table identifying the assigned disk unit for such server processor.

4. The method of claim 1, including the step of designating a one of the two or more server processors as a mount manager responsible for creating and maintaining a mount points table that identifies which of the disk units is assigned to which of the two or more server processors.

5. The method of claim 1, including the step of detecting a failure of the assigned disk unit by a one of the two or more server processors to send a message to the mount manager for assignment of a replacement disk unit.

6. The method of claim 1 including the steps of:

providing the master disk unit with a number of disk portions;

providing each of the other of the plurality of disk units with corresponding disk portions;

maintaining at the storage system a Data Status Bitmap Table to identify whether data written to a one of the disk portions of the master disk data has been copied to the other of the plurality of disk units.

7. The method of claim 6, wherein the writing step includes modifying the Data Status Bitmap Table to indicate that data written to the master disk unit has not been copied to the other of the plurality of disk units.

8. The method of claim 7, wherein the copying step includes changing the Data Status Bitmap Table for each of the other of the plurality of disk units to which the data is copied that the data has been copied thereto.

9. A data processing system, including:

a number of processors;

a storage system having a plurality of storage units, including a master storage unit, the storage system being communicatively coupled to the number of processors;

the number of processors including a mount manager operating to assign to each of the number of processors a corresponding one of the plurality of storage units;

the storage system including a disk controller operable to write data from the number of processors to the master disk unit and then copy the data to each of the corresponding ones of the plurality of storage units, each of the number of processors reading data from the assigned one of the plurality of storage units.

10. The data processing system of claim 9, including a bus structure for communicatively connecting the mount manager to the other of the number of processors.

11. The data processing system of claim 9, wherein the master storage unit includes a first storage space for storing a predetermined amount of data, and each of the other of the plurality of storage units having a second storage space for storing at least the predetermined amount of data.

12. The data processing system of claim 9, wherein each of the plurality of storage units are physical disk elements.

13. The data processing system of claim 9, including a data structure accessible to the disk controller for identifying when data is written to the master storage unit and copied to the other of the storage units.

14. The data processing system of claim 13, wherein the disk controller operates to consult the data structure when an I/O read request is received from a one of the processors to read data from the corresponding one of the storage units if the data is written to the master storage unit and copied to the one storage unit, otherwise to read the data from the master storage unit.

15. A data storage system operable to store and retrieve data in response to I/O write and read requests, respectively, from a plurality of processor elements, including:

a master storage unit and a number of mirrored storage units;

a controller that receives the I/O write requests to write data to the master storage unit and to each of the mirrored storage units;

there being an assignment of at least each of the mirrored storage units to corresponding ones of the plurality of processor elements, the controller receiving an I/O read request from a one of the processor elements to read data from the corresponding one of the storage units assigned to such processor element.

16. The data storage of claim 15, including a data structure accessible to the disk controller for identifying when data is written to the master storage unit and copied to the other of the storage units, the disk controller receiving the I/O read request to read data from the assigned mirrored storage unit if data written to the master storage unit has been copied to the assigned mirrored storage unit, else to read data from the master storage unit.

17. The data storage of claim 16 wherein the master and mirrored storage units are disk storage units.