Scalable object storage using multicast transport

• US 9,479,587 B2
• Filed: 06/11/2014
• Issued: 10/25/2016
• Est. Priority Date: 01/23/2013
• Status: Active Grant

• Current Assignee: Nexenta by DDN, Inc.
• Original Assignee: Nexenta Systems Incorporated (DataDirect Networks Incorporated)
• Inventors: Aizman, Alexander, Bestler, Caitlin, Novak, Robert E.
• Primary Examiner(s): Whipple, Brian P
• Application Number:
  US14/302,298
• Publication Number:
  US 20140304357A1
• Time in Patent Office: 867 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: H04L 12/1863 comprising mechanisms for improved reliability, e.g. status reports arrangements for detecting or preventing errors by carrying supervisory signal the return channel H04L1/16
• CPC Class Codes: H04L 49/00 Packet switching elements
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
• CPC Class Codes: H04L 69/164 Adaptation or special uses of UDP protocol

Abstract

Embodiments disclosed herein provide a scalable multicast transport. The multicast transport protocol provides effectively reliable multicast delivery while avoiding the overhead associated with point-to-point protocols. Additional embodiments disclosed herein relate to a scalable object storage system that uses a multicast transport. The object storage system assigns responsibility for providing storage services for a chunk to a negotiating group of storage servers in the cluster using a shared and distributed hash allocation table. The object storage system dynamically determines a rendezvous group of storage servers in the cluster to store the chunk using the multicast transport. Other embodiments, aspects and features are also disclosed.

Claims

1. A computer-implemented method of distributing chunks in a cluster of storage servers, the method comprising:

assigning, by the storage servers in the cluster, responsibility for providing storage services for a chunk to a first subset of storage servers in the cluster using a shared and distributed hash allocation table; and

dynamically determining, by at least one of the storage servers in the cluster, a second subset of storage servers in the cluster to store the chunk using a multicast transport service.

2. The method of claim 1, wherein the second subset of storage servers is a subset of the first subset of storage servers.

3. The method of claim 2, wherein the storage servers in the first subset of storage servers communicate with each other using multicast messaging to determine the second subset of storage servers.

4. The method of claim 1, wherein the first subset of storage servers is determined by a storage server in the cluster by mapping a hash identifier associated with the chunk to a tuple specifying a multicast address to reach the first subset of storage servers and enumerating storage servers in the first subset.

5. The method of claim 1, further comprising:

multicasting a get request to the first subset of storage servers for the chunk, wherein the get request is extended to specify a relative chunk position in an overall delivery of chunks referenced in a first retrieved chunk.

6. The method of claim 1, further comprising:

editing the hash allocation table by multicasting a versioned edit.

7. The method of claim 6, wherein the versioned edit includes a version number comprised of a timestamp and a source identifier of a storage server submitting the versioned edit.

8. A cluster of storage servers comprising:

a plurality of storage servers of the cluster;

a network of non-blocking switches communicatively interconnecting the plurality of storage servers of the cluster; and

an initiating client comprising a client system that communicatively interconnects to the network,

wherein the plurality of storage servers of the cluster assign responsibility for providing storage services for a chunk to a first subset of storage servers in the cluster using a shared and distributed hash allocation table, and

wherein at least one of the plurality of storage servers in the cluster dynamically determines a second subset of storage servers in the cluster to store the chunk using a multicast transport service.

9. The cluster of storage servers of claim 8, wherein the second subset of storage servers is a subset of the first subset of storage servers.

10. The cluster of storage servers of claim 9, wherein the storage servers in the first subset of storage servers communicate with each other using multicast messaging to determine the second subset of storage servers.

11. The cluster of storage servers of claim 8, wherein the first subset of storage servers is determined by a storage server in the cluster by mapping a hash identifier associated with the chunk to a tuple specifying a multicast address to reach the first subset of storage servers and enumerating storage servers in the first subset.

12. The cluster of storage servers of claim 8, wherein a get request is multicast to the first subset of storage servers for the chunk, wherein the get request is extended to specify a relative chunk position in an overall delivery of chunks referenced in a first retrieved chunk.

13. The cluster of storage servers of claim 8, wherein the hash allocation table is edited by multicasting a versioned edit.

14. The cluster of storage servers of claim 13, wherein the versioned edit includes a version number comprised of a timestamp and a source identifier of a storage server submitting the versioned edit.