System and method for concurrent distributed snapshot management

• US 6,895,415 B1
• Filed: 08/18/1999
• Issued: 05/17/2005
• Est. Priority Date: 08/18/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Rezaul Islam, Shah Mohammed, Sarkar, Prasenjit, Krishnamurthy, Vikram Harakere
• Primary Examiner(s): Robinson, Greta
• Assistant Examiner(s): Le, Debbie M.
• Application Number:
  US09/376,832
• Time in Patent Office: 2,099 Days
• Field of Search: 707200-204, 714/6, 714/1.21, 711/112, 711/162
• US Class Current: 1/1
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0617 in relation to availability
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD
• CPC Class Codes: Y10S 707/99953 Recoverability
• CPC Class Codes: Y10S 707/99955 Archiving or backup

Abstract

A computer-implemented method for managing data snapshots among storage volumes to facilitate transitive, distributed snapshot operations such that a read request or write request can be made to any volume in a snapshot includes establishing respective snapshot tables for the volumes. Each table can correlate, for each chunk of data, a source volume and a target volume. In response to a read request for chunks in one the volumes, the respective snapshot table is accessed, and the read directed to the source volume indicated in the snapshot table. Likewise, in response to a write request for chunks in a storage volume, the table associated with the volume is accessed, and the write is directed to the target volume indicated in the table. For each chunk, the source volume is the same across all tables, such that a read or write involves no more than a single storage volume.

Claims

1. A computer-implemented method for managing data snapshots among at least first and second storage volumes, comprising:

invoking a snapshot relationship between the volumes;

in response to the invoking act, establishing a logical copy of data from the first storage volume on the second storage volume;

establishing respective first and second snapshot tables for the volumes, each table including a source volume column and a target volume column;

in response to a read request for one or more read chunks in a “

read”

one of the volumes, accessing the snapshot table associated with the “

read”

volume,

either one of the first and second storage volumes being selectable as the read volume;

directing the read to the source volume as indicated for the read chunks in the source column of the snapshot table associated with the “

read”

volume;

in response to a write request for one or more write chunks in a “

write”

one of the volumes, accessing the table associated with the “

write”

volume,

either one of the first and second storage volumes being selectable as the write volume; and

directing the write to the target volume as indicated for the write chunks in the target column of the table associated with the “

write”

volume.

2. The method of claim 1, wherein the chunks are of equal size with each other.

3. The method of claim 1, wherein for each chunk, the source volume is the same across all tables, such that a at least one of:

a read, or a write, involves no more than a single storage volume.

4. The method of claim 3, further comprising at least a third storage volume with associated third snapshot table.

5. A computer program product, comprising:

a computer program storage device;

computer-readable instructions on the storage device for causing a computer to undertake method acts to facilitate transitive, distributed snapshot operations, the method acts comprising;

establishing a snapshot relationship wherein data chunks on a source volume are to be logically mirrored on a target volume;

for each volume, establishing an associated table that can list, for each chunk, a source volume and a target volume;

permitting subsequent I/O requests to either the source or target volume; and

in response to the I/O events, accessing the table of the source or target volume and directing the I/O event according to one or more entries therein.

6. The computer program product of claim 5, wherein the method acts embodied by the instructions include:

in response to the act of establishing a snapshot relationship, establishing a logical copy of data from the source volume on the target volume;

in response to a read request for one or more read chunks in a “

read”

one of the volumes, accessing the snapshot table associated with the “

read”

volume;

directing the read to the source volume indicated for the read chunks in the snapshot table associated with the “

read”

volume;

in response to a write request for one or more write chunks in a “

write”

one of the volumes, accessing the table associated with the “

write”

volume; and

directing the write to the target volume indicated for the write chunks in the table associated with the “

write”

volume.

7. The computer program product of claim 6, wherein the chunks are of equal size with each other.

8. The computer program product of claim 6, wherein for each chunk, the source volume is the same across all tables, such that a at least one of:

a read, or a write, involves no more than a single storage volume.

9. The computer program product of claim 8, wherein the method steps embodied by the instructions include changing entries in the tables in response to physical copying of data between the volumes as a result of a snapshot invocation.