Phased rollout of version upgrades in web-based business information systems

• US 7,970,901 B2
• Filed: 06/04/2009
• Issued: 06/28/2011
• Est. Priority Date: 07/12/2004
• Status: Active Grant

• Current Assignee: NetSuite Inc. (Oracle Corporation)
• Original Assignee: NetSuite Inc. (Oracle Corporation)
• Inventors: Lipscomb, David, Blum, Christopher T., Rice, Theodore R.
• Primary Examiner(s): Dharia; Rupal D
• Assistant Examiner(s): HENRY, MARIEGEORGES A
• Application Number:
  US12/478,590
• Publication Number:
  US 20090265425A1
• Time in Patent Office: 754 Days
• Field of Search: 709/217, 709/219, 709/225, 707/101, 717/170
• US Class Current: 709/225
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 99/00 Subject matter not provided for in other groups of this subclass
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

A method, system, computer program product, and related business methods for phased rollout of an upgraded version of a web-based business information system are described. URLs associated with client requests remain the same even where the target subscriber accounts have been upgraded. Client requests for both upgraded and non-upgraded accounts are received by a common set of web servers. Client requests representing external entry points are trapped, and the receiving web server accesses a login router database to determine whether the target subscriber account is upgraded or non-upgraded, and then transfers the request to an appropriately-versioned application server. Application server sets are mapped to database sets according to system version, allowing the upgraded system database schema to be substantially different than the non-upgraded database schema. An early adopter program business method leveraging the abilities of such web-based business information system is also described. Applications to environments other than phased rollout are described, such as URL-agnostic differentiation and routing of client requests according to class-of-service of the target subscribing account.

Claims

1. A method for provisioning phased rollout in a web-based business information system to facilitate upgrade from a current version to an upgraded version, the web-based business information system comprising a plurality of web servers for receiving client requests, a plurality of application servers for receiving the client requests from the web servers and processing the client requests, and a plurality of databases for access by the application servers during said processing, the method comprising:

identifying a set of subscriber accounts to be upgraded from the current version to the upgraded version;

migrating business data corresponding to said identified subscriber accounts from a first set of the databases configured in accordance with the current version to a second set of the databases configured in accordance with the upgraded version, the first set of the databases maintaining a first business data schema distinct from a second business data schema maintained by the second set of the databases;

modifying a first set of said plurality of application servers to process client requests according to said upgraded version in conjunction with said second set of the databases, and leaving a second set of said plurality of application servers to process client requests according to said current version in conjunction with said first set of the databases;

programming said web servers to trap client requests according to a predetermined set of external entry point URLs into the web-based business information system;

programming said web servers (i) to determine, for each trapped client request, whether said trapped client request is associated with the second business data schema, (ii) to forward said trapped client request to one of said first set of application servers if associated with the second business data schema, and (iii) to forward said trapped client request to one of said second set of application servers if not associated with the second business data schema; and

programming said web servers (i) to derive, based on a cookie or cookie-like data item contained in each non-trapped client request, whether said non-trapped client request is associated with the second business data schema, (ii) to forward said non-trapped client request to one of said first set of application servers if associated with the second business data schema, and (iii) to forward said non-trapped client request to one of said second set of application servers if not associated with the second business data schema.

2. The method of claim 1, wherein each of said external entry point URLs is characterized in that, if submitted by a requesting client to the web-based business information system in the absence of a previously established session between that requesting client and the web-based business information system, a login process therein would be instantiated.

3. The method of claim 1, further comprising programming each of said first and second sets of application servers to provide a cookie or cookie-like data item in conjunction with each response to a client request, said cookie or cookie-like data item identifying whether the responding application server is a member of said first or second set of application servers.

4. The method of claim 3, further comprising:

modifying a login router database to map said identified subscriber accounts into said second set of the databases; and

modifying said login router database to identify said second set of the databases as being associated with the second business data schema;

wherein said login router database is used by said web servers to perform said determining for said trapped client requests.

5. The method of claim 4, wherein each web server stores a separate copy of said login router database.

6. The method of claim 4, wherein said login router database comprises a first portion stored at a login router database server distinct from said web servers, and wherein each web server stores a separate copy of a second portion of said login router database.

7. The method of claim 6, wherein each web server further caches information derived from said first login router database portion when retrieved from said login router database server.