Snapshot based replication

• US 8,627,024 B2
• Filed: 11/30/2010
• Issued: 01/07/2014
• Est. Priority Date: 11/30/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Gupta, Karan, Haskin, Roger L, Sarkar, Prasenjit, Subhraveti, Dinesh K.
• Primary Examiner(s): YU, JAE UN
• Application Number:
  US12/957,285
• Publication Number:
  US 20120137094A1
• Time in Patent Office: 1,134 Days
• Field of Search: 711/162, 711/161, 711/169, 714/6, 714/6.231
• US Class Current: 711/162
• CPC Class Codes: G06F 16/184 implemented as replicated file system

Abstract

Embodiments of the invention relate to data replication and block allocation in a file system to support write transactions. Regions in a cluster file system are defined to support a block allocation. Blocks in the defined regions are allocated to support the data replication. A pipeline manager is provided to schedule population of the blocks in the allocated region(s) based upon network characteristics.

Claims

1. A system comprising:

a cluster file system having server nodes in communication across a network, the cluster including a first data storage local to a first server node, a second data storage local to a second server node, and a third data storage local to a third server node;

a write manager to write a first copy of data from a client machine to the first data storage local to the first server node;

an allocation manager in communication with the write manager, the allocation manager to allocate storage for a second copy of the data on the second data storage local to the second server node and to allocate storage for a third copy of the data on the third data storage local to the third server node in response to a pipeline manager replication of the first copy of data, wherein the allocation manager to create a second pointer from the allocated storage of the second server node to the first copy of data and to create a third pointer from the allocated storage of the third server node to the first copy of the data, and further comprising the allocation manager to remove the second pointer following completion of population of the second data storage local to the second server node with the first copy of the data, and to remove the third pointer following completion of population of the third data storage local to the third server node with the first copy of the data; and

the pipeline manager in communication with the allocation manager, the pipeline manager to schedule population of the second data storage and the third data storage with the first copy of the data based upon network characteristics.

2. The system of claim 1, further comprising the second pointer to re-direct a second read request from the second server node to the first node, and the third pointer to re-direct a third read request from the third server node to the first server node.

3. The system of claim 2, further comprising a read manager to self direct a read request to the second node following removal of the second pointer, and to self direct a read request to the third node following removal of the third pointer.

4. The system of claim 1, wherein scheduling by the pipeline manager is opportunistic with respect to network bandwidth availability.

5. A computer program product, the computer program product comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising:

computer readable program code configured to write a first copy of data on a first data storage local to a first server node;

computer readable program code configured to pipeline replication of the first copy of data on the first server node, including allocating storage for a second copy of the data on a second data storage local to a second server node and allocating storage for a third copy of the data on a third data storage local to a third server node, wherein the computer readable program code to allocate storage further comprising computer readable program code configured to create a second pointer from the allocated storage of the second server node to the first copy of data and to create a third pointer from the allocated storage of the third server node to the first copy of data, and further comprising computer readable program code configured to remove the second pointer following completion of populating the second data storage local to the second server node with the first copy of the data, and computer readable program code to remove the third pointer following completion of populating the third data storage local to the third server node with the first copy of the data; and

computer readable program code configured to schedule populating the allocated storage on the second data storage and the allocated storage on the third data storage with the first copy of the data based upon network characteristics.

6. The computer program product of claim 5, further comprising computer readable program code configured to have the second pointer re-direct a second read request from the second server node to the first node, and the third pointer re-direct a third read request from the third server node to the first server node.

7. The computer program product of claim 6, further comprising computer readable program code configured to self direct a read request to the second node following removal of the second pointer, and to self direct a read request to the third node following removal of the third pointer.

8. The computer program product of claim 5, wherein the computer readable program code to schedule populating the second data storage and the third data storage with the first copy of the data is opportunistic with respect to network bandwidth availability.