On-demand capacity management

• US 7,925,785 B2
• Filed: 06/27/2008
• Issued: 04/12/2011
• Est. Priority Date: 06/27/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Brown, Michael E., Ruia, Anil K., Yoo, Won Suk
• Primary Examiner(s): WON, MICHAEL YOUNG
• Application Number:
  US12/163,934
• Publication Number:
  US 20090327459A1
• Time in Patent Office: 1,019 Days
• Field of Search: 709/220, 709/226, 709/242, 709/201, 709/217, 709/227, 709/232, 709/238, 709/252, 709/223
• US Class Current: 709/241
• CPC Class Codes: H04L 45/00 Routing or path finding of packets in data switching networks routing or path finding in wireless networks H04W40/00
• CPC Class Codes: H04L 45/22 Alternate routing
• CPC Class Codes: H04L 47/10 Flow control; Congestion control

Abstract

Dynamically upsizing and/or downsizing a network farm in response to network demand. An application message router routes messages to the network farm. When the network farm approaches or is anticipated to be approaching capacity, a group of one or more servers may be added to the network farm. When the added server(s) are capable of participating in the network farm, the application message router is triggered to route also to the added servers. When the network farm has excess capacity, a group of one or more servers may be dropped from the network farm. This may be accomplished by triggering the application message router to no longer route messages to the removed servers. The removed servers may be either immediately or gracefully removed from service.

Claims

1. A method for dynamically upsizing a network farm of one or more servers with at least one additional server in response to network demand, the method comprising:

initially routing incoming application messages to an initial set of one or more server(s) using an application message router that uses configurable routing settings;

identifying a routing policy for the network farm;

statistically tracking information corresponding to the initially routed application messages;

determining from the tracked information that the one or more servers are reaching capacity;

in response to determining that the one or more servers are reaching capacity,

adding an additional server to the network farm; and

modifying the configurable intra-farm routing settings such that subsequent incoming messages are routed to a subsequent set of servers comprising the initial set of server(s) and the additional server,

wherein modifying the configurable routing settings is performed using an application program interface that provides a first interface that permits for immediate rerouting of incoming messages away from one or more servers of the network farm that are not in the diminished set and provides a second interface that permits rerouting of incoming messages with at least temporary continuance of routing of incoming messages that correspond to pre-existing stateful sessions to their pre-existing servers even if those pre-existing servers are no longer in the diminished set of server(s).

2. A method in accordance with claim 1, wherein adding the additional server to the network farm comprises:

making the additional server physically available on the network farm to receive incoming application messages; and

configuring the additional server to be capable of responding to at least some of the incoming application messages.

3. A method in accordance with claim 1,

wherein adding the additional server is performed in response to determining from the tracked information that the one or more servers are reaching capacity.

4. A method in accordance with claim 3, wherein statistically tracking information is performed with the aid of a second first application program interface that reports regarding one or more performance parameters of the network farm.

5. A method in accordance with claim 4, wherein one of the one or more performance parameters comprises an indication of a response time for a request or the amount of work the network farm does to respond to the request.

6. A method in accordance with claim 4, wherein the first and second application program interfaces are part of the same program.

7. A method in accordance with claim 4, wherein the first and second application program interfaces and a routing application that manages the configurable routing rules are all part of the same program.

8. A method in accordance with claim 1, wherein the subsequent set of servers is a first subsequent set of servers, the method further comprising the following after modifying the configurable routing settings such that incoming messages are routed to the first subsequent set of servers:

removing one of the servers from the network farm; and

modifying the configurable routing settings such that the incoming messages are routed to a second subsequent set of one or more servers comprising the first subsequent set of servers minus the removed one or more servers.

9. A method in accordance with claim 8, further comprising the following before removing one of the servers from the network farm:

detecting that the second subsequent set of servers is now capable of meeting current demand without the one or more servers to be removed in the act of removing.

10. A method for dynamically downsizing by one or more servers a network farm comprising a plurality of servers in response to diminished demand, the method comprising:

identifying a routing policy for the network farm;

statistically tracking information corresponding to demand upon a first one or more servers;

determining from the tracked information that the network farm is capable of meeting current demand without the first one or more servers; and

modifying configurable routing settings such that subsequent incoming messages are routed to a diminished set of second one or more servers of the network farm in response to determining from the tracked information that the network farm is capable of meeting current demand without the first one or more servers, the diminished set not including the first one or more servers,

wherein modifying the configurable routing settings is performed using an application program interface provides a first interface that permits for immediate rerouting of incoming messages away from one or more servers of the network farm that are not in the diminished set and provides a second interface that permits rerouting of incoming messages with at least temporary continuance of routing of incoming messages that correspond to pre-existing stateful sessions to their pre-existing servers even if those pre-existing servers are no longer in the diminished set of server(s).

11. A method in accordance with claim 10, further comprising:

removing one of the servers from the network farm.

12. A method in accordance with claim 11, wherein the removing the one or more servers from the network farm comprises:

making the one or more servers unavailable for incoming messages have a new session or for incoming messages corresponding to a stateless session; and

keeping the one or more servers available temporarily to handling incoming messages of a preexisting stateful session.

13. A computer program product comprising one or more computer readable devices having encoded thereon one or more computer-executable instructions that, when executed by one or more processors of a computing system, cause the computing system to provide an application program interface, the application program interface comprising:

a reporting interface that is configured to report regarding one or more performance parameters of a network farm to thereby allow a user or other entity to determine when a network farm is reaching capacity, the reported performance parameters having been identified by identifying a routing policy for the network farm and statistically tracking information corresponding to the one or more performance parameters;

an augmentation interface that is configured to allow for immediate rerouting of incoming application networks to an augmented network farm in the case where one or more servers are added to the network farm; and

a diminution interface that is configured to allow for immediate rerouting of an incoming application message to a diminished network farm in the case where one or more servers are removed from the network farm, wherein the diminution interface comprises;

an immediate diminution interface that causes all incoming application messages to be routed to the diminished network farm and causes all pending requests at the removed server(s) to be terminated;

a graceful diminution interface that causes all incoming application messages to be routed to the diminished network farm, but allows at least some of the pending requests at the removed server(s) to be continued to completion; and

a graceful diminution interface that allows some of the removed servers to temporarily continue serving incoming application messages that correspond to pre\-existing stateful sessions that they had been previously serving.