Method and apparatus for web based storage on demand
Rapid demanding for storage capacity at internet era requires a much flexible and powerful storage infrastructure. Present invention disclosed a type of storage system based a model of centrally controlled distributed scalable virtual machine. In this model, one or more service pools including virtual storage service pool and application service pools can be automatically created to meet the demands for more storage capacity from various applications. Specially this model provide a solid foundation for distributing storage volumes for supporting storage on-demand and sharing with exceptional management capabilities. In addition, this model provides a flexible fault recovery topology beyond the traditional recovery plan.
1. A computing system implementing a topology for fault recovery, the computing system comprising:
a plurality of storage servers and a plurality of hosts interconnected across a network infrastructure;
wherein the storage servers are organized to have a plurality of mirrored pairs, each of which operable delivering storage service to one of the hosts, and
wherein at least a first one of the storage servers is configured as a first spare storage server for delivering fault replacement service to a first one of the mirrored pairs whenever a fault occurred to one of mirrored storage servers in the first one of the mirrored pairs; and
wherein each of the hosts is configured for access to at least one of the mirrored pairs, and
wherein the hosts are configured with at least a first spare host and two or more non-spare hosts among the hosts, the first spare host being configured for delivering fault replacement service to the two or more non-spare hosts whenever a fault occur to a first one of the two or more non-spare hosts while the first one of the two or more non-spare hosts access to storage service delivered by the first one of the mirrored pairs.
2. The computing system of claim 1 further comprising:
constructing the network infrastructure to be operable across multiple network domains for implementing the fault recovery cross a distance beyond hundreds miles long.
3. The computing system of claim 1 , wherein the network infrastructure comprises at least a first network for the storage servers and a second network for the hosts, and comprises data communication links between the first network and the second network.
4. The computing system of claim 3 , wherein the first networks is one of a corporate storage network, corporate Intranet, Internet, or wide are network (“WAN”), and the second network is one of a corporate Intranet, Internet, or WAN.
5. The computing system of claim 1 , wherein the mirrored storage servers in each of the mirrored pairs comprise a primary storage sever and a secondly storage server, the primary storage server storing data written by one of the hosts while the secondly storage server keeps and maintains a copy of the data which mirrors the data stored in the primary storage server.
6. The computing system of claim 5 , wherein the secondly storage server further delivers continued storage service to the one of the hosts based on the mirrored copy of the data whenever a fault occur to the primary storage server.
7. The computing system of claim 5 , wherein the first spare storage server is configured to be operable to deliver the fault replacement service to the primary storage server or to the secondly storage server whenever a fault occur to the primary or secondly storage server in the first one of the mirrored pairs.
8. The computing system of claim 1 , wherein the first spare storage server is configured to be operable to deliver the fault replacement service to a second one of the mirrored pairs.
9. The computing system of claim 1 , further comprising: configuring a second spare host among the hosts for delivering the fault replacement service to another two or more non-spare hosts among the hosts.
10. The computing system of claim 1 , further comprising:
configuring a second one of the storage servers as a second spare storage server for delivering fault replacement service to two or more of the mirrored pairs.
11. A method for implementing a fault recovery topology of a computing system comprising:
constructing a network infrastructure to comprise a first network of a plurality of storage servers and a second network of a plurality of hosts and comprise data links between the hosts and the storage servers;
organizing the storage servers into a plurality of mirrored pairs each of which operable delivering storage service to one of the hosts, and configuring
at least a first one of the storage servers as a first spare storage server for delivering fault replacement service to a first one of the mirrored pairs whenever a fault occurred to one of mirrored storage servers in the first one of the mirrored pairs; and
configuring each of the hosts access to at least one of the mirrored pairs, and configuring at least a first spare host and two or more non-spare hosts among the hosts, the first spare host delivering fault replacement service for the two or more non-spare hosts whenever a fault occur to a first one of the two or more non-spare hosts while the first one of two or more non-spare hosts access to the storage service delivered by the first one of the mirrored pairs.
12. The method of claim 11 further comprising:
constructing the network infrastructure for implementing the fault recovery cross a distance beyond hundred miles long.
13. The method of claim 11 , wherein the network infrastructure comprises data communication links between the first network and the second network.
14. The method system of claim 13 , wherein the first networks is one of a corporate storage network, corporate Intranet, Internet, or wide are network (“WAN”), and the second network is one of a corporate Intranet, Internet, or WAN.
15. The method of claim 11 , wherein said mirrored storage servers in each of the mirrored pairs comprise a primary storage server and a secondly storage server, the primary storage server storing data written by one of the hosts while the secondly storage server keeps and maintains a copy of the data which mirrors the data stored in the primary storage server.
16. The method of claim 15 , wherein the first spare storage server is configured to be operable to deliver the fault replacement service to the primary storage server or to the secondly storage server whenever a fault occur to the primary or secondly storage server.
17. The method of claim 11 , wherein the first spare storage server is configured operable to deliver the fault replacement service to a second one of mirrored pairs.
18. The method of claim 11 , comprising: configuring a second one of the storage servers as a second spare storage server for delivering fault replacement service to two or more of the mirrored pairs.
19. The method of claim 11 , comprising: configuring a second spare host among the hosts for delivering fault replacement service to another two or more non-spare hosts among the hosts.