Problem statement
The company's main office and a remote facility (e.g., a branch office) have a telephony system based on ECSS-10 Softswitch. It is necessary to ensure that the telephony system continues to operate in the event of a single PBX failure. For this purpose, geographic redundancy (geo-redundancy) is organized between the sites. In normal mode, subscribers work through the local PBX. In case of its failure, they are re-registered on the backup, and after restoration of operability, they return to the main one. For offices with SMG transit registration, it will also switch to the backup higher-level server and, only if it is unavailable, serve subscribers on its own.
Figure1. Geographic redundancy
The principle of geographic redundancy
Geographic redundancy operates at the domain level. Thus, one ECSS-10 site can be primary for one domain and redundant for another. In this guide, an ECSS-10 site refers to a Softswitch installation deployed in a cluster or on a single server. Geo-reserve can be created between sites with Softswitch clusters, between single-node installations, or between a single-node installation and a cluster.
Devices registering with Softswitch must be configured to switch between the primary and backup sites in Homing mode.
In Homing mode, the main site is constantly polled. If it becomes unavailable, a switch to the backup site occurs. Monitoring of the main site's availability continues. When the main site becomes available, a return to the main site occurs.
Configuration
A complete copy of the main domain is created on the backup site. In case of complete loss of the main site, the backup site can be assigned by the replication master and a new backup site can be configured.
/domain/<domain_name>/properties/set replica_type master
It is also possible to change the replication direction: make the main site a backup site and the backup site a main site. However, it should be noted that replication is performed periodically, and changes made after the last replication may be lost during such a switch.
Work details
During a failure of the primary site, call recording and CDR will be performed on the backup site.