Data storage system

• US 20070226224A1
• Filed: 03/08/2006
• Published: 09/27/2007
• Est. Priority Date: 03/08/2006
• Status: Abandoned Application

• Current Assignee: Harmonic Incorporated
• Original Assignee: Omneon Video Networks
• Inventors: Craig, Donald, Howe, John, Mitaru, Alexandru, Wanigasekara-Mohotti, Don, Davis, Christopher
• Application Number:
  US11/371,392
• Publication Number:
  US 20070226224A1
• Time in Patent Office: Days
• US Class Current: 1/1
• 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]

Abstract

The data storage system has metadata server machines, content server machines arranged in groups where each group has a respective one or more of the content server machines, and a system interconnect to which the metadata and content server machines are communicatively coupled. A distributed file system is to be executed in the metadata server machines, the content server machines, and the client machine, to hide complexity of the system from a number of client machine users. The content server machines store slices of the files at locations indicated by the metadata. Other embodiments are also described and claimed.

Claims

1. A data storage system that can be accessed from a client machine, comprising:

a plurality of metadata server machines each to store metadata for a plurality of files that are stored in the system;

a plurality of content server machines arranged as a plurality of groups, each group having a respective one or more of the content server machines, the content server machines to store slices of said files at locations indicated by the metadata;

a system interconnect to which the metadata and content server machines are communicatively coupled; and

a distributed file system to be executed in the metadata server machines, the content server machines, and the client machine,

the distributed file system to hide complexity of the system from a plurality of client machine users and spread each of the files over two or more of the content server machines and two or more of the groups.

2. The system of claim 1 wherein the server machines are grouped on the basis of having one or more common installation parameters, examples being, but not limited to, a) sharing the same power source, b) having the same hard disk drive model type, and c) connected to the same packet switch in the system interconnect.

3. The system of claim 2 wherein the number of groups and the membership of content server machines in groups can be varied dynamically while providing continuous client machine access to files stored in the file system.

4. The system of claim 3 wherein the files and their slice replicas are spread out over the content server machines, so that slice replicas are in as many groups as possible for a given number of replicas.

5. The system of claim 1 wherein the distributed file system accepts a client-specified number of acknowledgements for replicating the slices of a given file.

6. The system of claim 1 wherein the file system can accept:

a) a first acknowledgement level for which it will provide a client with acknowledgement of replication before the given file has been replicated a set of number of times, and

b) a second acknowledgement level for which it will provide a client with acknowledgement of replication only after the given file has been replicated the set of number of times.

7. The system of claim 1 wherein each of the slices of a given file is self-locating so that in the event metadata for the given file is lost, each of the slices can still identify the given file without the need to refer to the given file'"'"'s metadata.

9. The system of claim 1 wherein the system interconnect is split into two independent subnets such that the metadata servers and content servers are able to detect loss of connectivity in one subnet and continue operation by utilizing the remaining subnet.

8. A method for operating a data storage system, comprising:

receiving a first client request to create a file and responding with a file handle that indicates where a slice of the file is or will be stored in the system;

receiving a second client request including the file handle, to create slices of the file and responding with the identities of content servers that are members of different groups across groups; and

receiving a plurality of client requests that specify different replication factors for the file, and responding by changing the number and location of slice replicas in order to maintain the spread of slice replicas across groups.

10. An article of manufacture, comprising:

a machine-readable medium having instructions stored therein that when executed

a) issue a first client request to access a file and expect a file handle that indicates where a slice of the file is or will be stored in a data storage system, wherein the file is to be stored in the system as divided into a plurality of slices and the slices are to be replicated across groups in the system for redundancy,

b) issue a second client request including the file handle, to access the slices of the file, and

c) issue a plurality of client requests to the system that specify different replication factors for the file.