Method and system for fail-safe call survival

A method of providing a telecommunication service to a call, said call being established or to be established between at least two endpoints, the method comprising the steps of:

• a) providing a plurality of servers, at least one server being active to provide said service, at least one server waiting to be activated to provide said service;
• and
• b) providing a recovery device configured for:
• b1) determining unavailability of each server for providing said service by testing sanity of each of said plurality of servers and obtaining test results at a predetermined frequency;
• and
• b2) if an active server is determined to be unavailable, activating said at least one waiting server within a signaling timeout period to provide said service without interruption to said call and without involvement of said at least two endpoints;
• wherein each server comprises a watchdog program for performing independent functional testing of said service and resource utilization testing of each server and communicating test results to said recovery device, said independent functional testing being performed and results thereof being communicated at a frequency that is less than the signaling timeout period defined by said at least two endpoints of said call;
• and
• wherein said determining and activating are performed independent from said call.

A computer readable memory having recorded thereon statements and instructions for execution by a computer for providing a telecommunication service to a call, said call being established or to be established between at least two endpoints, said statements and instructions comprising:

• program code which, when executed, is operable to provide and configure a plurality of servers, at least one server being active to provide said service, at least one server waiting to be activated to provide said service;
• and
• program code which, when executed, is operable to determine an unavailability of each server for providing said service by testing sanity of each of said plurality of servers and obtaining test results at a predetermined frequency and activate said at least one waiting server, said determining and activating being independent from said call;
• wherein each server comprises a watchdog program for performing independent functional testing of said service and resource utilization testing of each server and communicating test results, said independent functional testing being performed and results thereof being communicated at a frequency that is less than a signaling timeout period defined by said at least two endpoints of said call;
• and
• wherein when an active server is determined to be unavailable, said at least one waiting server is activated within said signaling timeout period to provide said service without interruption to said call and without involvement of said at least two endpoints.