On Thursday 01 October 2015 07:35:06 Grant Bagdasarian wrote:
How do big carriers handle these kinds of scenarios where a sudden outage in a key component would cause calls to be dropped? Is this the achilles heel of voice? Is there special kind of software or hardware (or both) for handling these kinds of scenarios?
See http://www.ng-voice.com/redundancy-maxx/ for example.
Your specific concern: "Both our deployed SEMS (thanks to the Frafos Team) and our deployed RTPEngine (thanks to Carlos) use a Redis-Database for storing information on the current calls, however in both cases through a proprietary extension. The Redis- database is distributed among the two servers, so if we start the services on the passive node, all sessions can continue without service interuption."