VICIdial Cluster

A VICIdial cluster is a multi-server deployment where two or more VICIdial servers work together to share the call handling workload. Clustering enables horizontal scaling — adding more servers to handle more concurrent calls and agents — along with redundancy, so the operation can continue if a single server fails.

How It Works in VICIdial

A VICIdial cluster consists of multiple component server roles: web servers (hosting the admin panel and agent screens), telephony servers (running Asterisk for call handling), and a database server (MySQL, which stores all configuration, lead, and reporting data). In smaller clusters, multiple roles may share a single server; in larger deployments, each role runs on dedicated hardware.

Telephony servers communicate with each other via IAX2 trunks for inter-server call routing and audio transfer. When a call connects on one Asterisk server but the assigned agent is registered on a different server, the audio is bridged between servers over IAX2. The centralized MySQL database coordinates which server handles which calls, agents, and campaigns.

Database replication (MySQL master-slave or Galera cluster) provides database redundancy. Load balancing distributes agent connections and web traffic across multiple web servers.

Why It Matters

Single-server VICIdial deployments are limited to approximately 300-500 concurrent calls depending on hardware. Clustering removes this ceiling — large enterprises and BPOs run VICIdial clusters handling thousands of concurrent calls across dozens of servers. Clustering also provides business continuity; if one telephony server fails, agents on other servers continue operating.

Cluster design requires careful planning around network latency (inter-server IAX2 audio is sensitive to delay), database performance (the central MySQL instance becomes the bottleneck at scale), and server role allocation. See our VICIdial cluster guide for architecture planning and deployment strategies.