Load balancing method for exchanging data between multiple hosts and storage entities, in IP based storage area network

• US 7,464,156 B2
• Filed: 09/06/2002
• Issued: 12/09/2008
• Est. Priority Date: 09/07/2001
• Status: Expired due to Fees

• Current Assignee: OCZ Israel Limited (Kioxia Corporation)
• Original Assignee: Sanrad Ltd. (ZCO Liquidating Corporation)
• Inventors: Amir, Shai, Klein, Yaron
• Primary Examiner(s): Pwu, Jeffrey
• Assistant Examiner(s): SMARTH, GERALD A
• Application Number:
  US10/488,491
• Publication Number:
  US 20050010682A1
• Time in Patent Office: 2,286 Days
• Field of Search: 709/201, 709/223, 709/220, 709/230, 709/238, 709/225, 370/35, 370/254, 370/389
• US Class Current: 709/224
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/101 based on network conditions
• 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 9/40 Network security protocols

Abstract

The present invention achieves load balancing of activities on storage entities, operating in a storage area network (SAN) by assignment of address to each storage entity. The storage traffic is monitored and statistics are accumulated over an interval and then used to obtain the traffic pattern of each host-storage entity link. A statistical analysis is used to determine the optimal routing map from hosts to storage entity. Other heuristics including genetic algorithms many also be used. This mapping is set on the network by assigning the entities'"'"' addresses to the storage gateways, thus does not impose any process on the hosts or the disks.

Claims

1. A method for balancing storage traffic in a storage area network (SAN), including at least a plurality of host computers, at least one storage gateway, and a plurality of storage entities, wherein the host computers are coupled to the storage gateway through a network and the storage entities are coupled to the storage gateway through a communication means, said method comprising:

accumulating, over a time interval, statistical readings on the storage traffic for all routes in the SAN, wherein a route is established between each a host computer and a storage entity through the at least one gateway;

determining a routing map using the statistical readings and a cost function, wherein the cost function is min (w_variance*variance(GatewayTraffic)+w_change*change), where the change value is a number of changes from a previous routing map, w_variance is a weight assigned to the variance, the W_change is a weight assigned to the change value, and the GatewayTraffic is the statistical readings on the storage traffic;

said cost function is computed for all combinations of routes in the SAN; and

using the routing map to assign new routes between the plurality of host computers and the plurality of storage entities through the at least one storage gateway;

whereby the new routes allow to balance the storage traffic in the SAN.

2. The method of claim 1, wherein the routes for which statistical readings are accumulated are pre-assigned.

3. The method of claim 2, wherein the pre-assignment of routes is performed a-priori by a network manager.

4. The method of claim 2, wherein the pre-assignment of routes is performed by the steps of:

selecting a unique storage gateway from the at least one storage gateway for connecting between a host computer and a storage entity; and

assigning an Internet protocol (IP) address to the storage entity such that the host computer recognizes the storage entity by the IP address.

5. The method of claim 1, wherein the routing map includes a set of routes that minimizes the cost function.

6. The method of claim 1, wherein the routing map may be computed using a heuristic algorithm including at least one of:

a simulated annealing algorithm, a greedy algorithm, a economic optimization algorithm, and a search algorithm.

7. The method of claim 1, wherein the assignment of new routes is performed by a storage name server (SNS).

8. The method of claim 7, further comprising:

periodically querying, by a host computer, the SNS for an IP address of a storage entity; and

granting the IP address of the storage entity to the host computer according to the routing map.

9. The method of claim 1, wherein the network is an IP network.

10. The method of claim 1, wherein the storage entity is at least one of:

a physical disk, a virtual disk, a cluster of physical disks, a cluster of virtual disks, and a cluster of virtual disks and physical disks.