Scalable storage system with unique client assignment to storage server partitions

• US 6,564,252 B1
• Filed: 03/11/1999
• Issued: 05/13/2003
• Est. Priority Date: 03/11/1999
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: DeMello, Marco A., Hickman, Robert Chris, Sinha, Alok K.
• Primary Examiner(s): Maung, Zarni
• Application Number:
  US09/266,458
• Time in Patent Office: 1,524 Days
• Field of Search: 709/215, 709/214, 709/213, 709/216, 709/220
• US Class Current: 709/214
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1031 Controlling of the operation of servers by a load balancer, e.g. adding or removing servers that serve requests
• CPC Class Codes: H04L 67/1036 Load balancing of requests to servers for services different from user content provisioning, e.g. load balancing across domain name servers

Abstract

A massively scalable architecture for providing a highly reliable storage system that is capable of handling hundreds of millions of users and tens of billions of files. The system includes one or more clusters storing data elements that are received from a plurality of clients. Each cluster comprises a plurality of storage servers. A partition master assigns each client a unique partition within one of the clusters such that each data element is stored in the partition assigned to the client. The storage system can be easily scaled as the number of reads increases by adding individual storage servers to each storage cluster. In addition, the storage system easily scales to handle an increase in the number of writes, or as the number of files per cluster exceeds a predefined limit, by adding new storage clusters to the storage system.

Claims

1. A storage system for managing data elements in a distributed storage environment, the storage system comprising:

a plurality of distributed clusters storing data elements from a plurality of clients, wherein each cluster comprises a plurality of storage servers and wherein the data elements are not centrally stored; and

a partition master communicatively coupled to the clusters, wherein the partition master assigns each client a cluster from the plurality of clusters and a unique partition within the assigned cluster, and further wherein each data element is stored in the partition assigned to the client and is replicated across the plurality of storage servers within the respective cluster.

2. The storage system of claim 1, wherein each cluster includes a write master receiving the data elements from the clients and directing the storage servers of the respective storage cluster to store the received data elements.

3. The storage system of claim 1, wherein each cluster includes a cluster backup storing all the data elements stored by the storage servers of the respective cluster.

4. The storage system of claim 1 further comprising a storage manager communicatively coupled to the clusters, wherein the storage manager displays a user interface allowing a system administrator to redistribute the partitions within the clusters.

5. The storage system of claim 1 further comprising a storage manager communicatively coupled to the clusters for detecting when one of the storage servers fails.

6. The storage system of claim 1, wherein the clients are Internet users.

7. The storage system of claim 1, wherein the clients are applications executing on at least one web server.

8. The storage system of claim 7, wherein the applications conform to an Internet Server Application Programming Interface (ISAPI).

9. The storage system of claim 8 further including a storage access module executing one each of the web servers for communicating storage access requests to the storage clusters.

10. The storage system of claim 9, wherein each web server includes a cache for mapping the access requests to one of the partitions on the clusters.

11. A computing method for managing data elements in a distributed storage environment, wherein data elements are not centrally stored, the method comprising:

receiving a client request to store data elements;

assigning the client a unique partition within one of a plurality of storage clusters; and

storing the data elements in each of a plurality of servers within the storage cluster.

12. The method of claim 11, wherein storing the data elements includes storing the data elements within a cluster backup.

13. The method of claim 11 further including redistributing the partitions within the system.

14. The method of claim 13, wherein redistributing the partitions is performed in response to a command received from a system administrator.

15. The method of claim 13, wherein redistributing the partitions includes sensing when a load on one of the clusters exceeds a predetermined threshold.

16. The method of claim 11 further including detecting when one of the storage servers fails.

17. The method of claim 11, wherein receiving the requests includes receiving requests from Internet users.

18. The method of claim 11, wherein receiving the requests includes receiving requests from applications executing on at least one web server.

19. The method of claim 18, wherein storing the data elements includes accessing a cache to map the access requests to one of the partitions on the clusters.

20. In a storage system for managing data elements in a distributed storage environment, wherein data elements are not centrally stored, a computer-readable medium having computer-executable instructions for causing a system to perform a computing method for a managing data elements, the computing method comprising:

receiving a client request to store data elements;

assigning the client a unique partition within one of a plurality of distributed storage clusters; and

storing the data elements in each of a plurality of servers within the assigned storage cluster.

21. A computing system for managing data elements in a distributed storage environment, the computing system comprising:

a plurality of webs servers receiving client requests to store data elements;

a plurality of distributed clusters storing the data elements received from the clients, wherein each cluster comprises a plurality of storage servers and wherein data elements are not centrally stored;

a partition master communicatively coupled to the clusters, wherein the partition master maintains partition information mapping each client to a unique partition within one of the clusters; and

a storage access module executing on each of the web servers, wherein for each client request the storage access module communicates the request to the storage clusters as a function of the partition information.

22. The system of claim 21, wherein the data elements stored within each cluster are replicated across the plurality of storage servers within the respective cluster.