Pan W Wang

US Patent:

8447881, May 21, 2013

Filed:

Sep 2, 2008

Appl. No.:

12/203024

Inventors:

Kartik Paramasivam - Redmond WA,
Stuart J. Langley - Redmond WA,
Vinod Shanbhag - Redmond WA,
Pan Wang - Bellevue WA,
Ramasubramaniam Poornalingam - Redmond WA,
Brent Schmaltz - Redmond WA,
Jayteerth Katti - Redmond WA,
Govindaraj Ramanathan - Redmond WA,

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 15/173

US Classification:

709238, 709203, 718105

Abstract:

The present invention extends to methods, systems, and computer program products for load balancing for services. Embodiments of the invention facilitate load balancing between instances of a service based on affinitization of messages, based on content of the message. For example, messages in the same session can be dispatched to the same service instance. A sequence or series of related messages associated with long running and/or stateful services are more likely to be dispatched to the same instance of the service. Thus, if the service instance has persisted client state, there is an increased likely of utilizing the persisted client state and not having another service instance recreate the client state.

US Patent:

2009032, Dec 31, 2009

Filed:

Jun 26, 2008

Appl. No.:

12/147155

Inventors:

Kartik Paramasivam - Redmond WA,
Pan Wang - Bellevue WA,

Assignee:

MICROSOFT CORPORATION - Redmond WA

International Classification:

G06F 9/46

US Classification:

718105

Abstract:

A system for efficiently distributing messages to a server farm uses a hashing function and a map-based function, or combinations thereof, to distribute messages associated with a processing request. In one implementation, for example, the hashing function has inputs of an identifier for each message in a processing request, and a list of available servers. Upon identifying that any of the servers is unavailable, or will soon be unavailable, the load balancing server uses an alternate map-based assignment function for new requests, and inputs each assignment into a server map. The load balancing server can then use the map or the hashing function, as appropriate, to direct messages to an operating server. Upon receiving an updated list of available servers, the load balancing server can switch back to the hashing function after the map is depleted, and use the updated server list as an argument.