why vicidial is slower

VICIdial slows down when the single box running it hits a resource ceiling, usually dialing rate, recording load, and database writes all competing for the same CPU and disk at once, or when the server's health degrades under load, a full disk, or a stalled process.

Common causes

Single-server resource limits If you're on one box, the limit is rarely agent count alone. It is the combination of dialing rate, recording load, and database writes all landing on the same CPU and disk. When any one of those starts to saturate, you feel it as slow agent screens, delayed lead delivery, or dropped audio. This typically shows up once a center grows past roughly 30 to 40 agents.

A degraded server If the server itself is struggling, you may see a DEGRADED health status. High system load means the box is under more demand than it can comfortably serve, and you may see slower response times, delayed dialing, or agents reporting lag before this becomes an outright failure. High load does not always mean too many calls; a runaway process or a slow database query can also pin load while call volume stays normal.

Disk filling up A full disk degrades everything quietly. Once the disk crosses a high threshold, the OS cannot write log files or temp files reliably. Call recordings fail silently. The database can start throwing write errors. The box is still reachable over SSH but its functions are impaired. This also applies to recordings specifically: the system does not automatically delete or expire recording files, so left unmanaged, the recording directory fills up and new recordings silently fail to write.

A stalled keepalive Dialing itself can stall without an obvious crash. A dead keepalive process is VICIdial's heartbeat of the auto-dialing engine. If one exits and does not restart, the dialer stops placing outbound calls even though the web admin still loads and agents can still log in. The symptom is a campaign that appears active but dials nothing.

What to check

• Run uptime and df -h on the server to rule out load and disk space, then verify that Asterisk is running and accepting connections, and check that the VICIdial keepalive processes are alive.
• If you're consistently maxing out one box, consider splitting the database and web layer onto their own machine or adding a second dialer, since a second common split is moving the database and web server onto their own dedicated machine so the dialers do nothing but make calls.

If none of this matches what you're seeing, let us know the specific symptom at /contact and we can dig further.