Resizing your dialer for a bigger campaign

When a Campaign grows beyond what your current server can handle comfortably, you can resize the underlying VPS from the VICIfast dashboard. The resize process is managed, near-instant, and takes a safety snapshot before it starts so you can roll back if needed.

This is a feature of managed VICIfast plans. If you are on the BYOI (bring your own infrastructure) plan, where VICIfast installs and manages VICIdial on a server you own, resize is not available because VICIfast does not control the hypervisor underneath your box.

Why campaign scale drives server size

VICIdial load scales with the number of Concurrent calls in flight at once. Each bridged call between a Carrier trunk and an Agent channel consumes CPU for audio mixing and memory for the Asterisk process. A predictive dialer running at 2.5 Lines per agent across twenty agents is placing fifty simultaneous outbound legs at peak. A box provisioned for ten agents at that ratio starts to show audio quality degradation and pacing delays well before it crashes.

The signal to resize is usually a combination of sustained high CPU on your real-time report, increasing call setup failures, and agents reporting choppy audio. If you are seeing those together, you have likely outgrown the current box class.

How the managed resize works

Resizing a Hetzner VPS in place means the server shuts down briefly, Hetzner migrates the disk image to a larger hardware class, and the server reboots on the new hardware. VICIfast wraps that in a state-machine transition with three guard steps.

• A Server snapshot is taken before the resize begins. If the resize fails or the VICIdial install does not come back clean after reboot, you have a point-in-time rollback target.
• The server status moves from ACTIVE to RESIZING. During this window, active calls on the box will drop. Schedule the resize during a low-traffic window for a live campaign.
• After the resize, VICIfast runs a health check. Once the box reports back HEALTHY, the status returns to ACTIVE and agents can log in to the upgraded server.

State transitions during a resize

stateDiagram-v2
  [*] --> ACTIVE
  ACTIVE --> RESIZING: Resize requested
  RESIZING --> ACTIVE: Health check passed on larger box
  RESIZING --> FAILED: Resize or health check failed
  FAILED --> ACTIVE: Restore from pre-resize snapshot

The entire cycle from ACTIVE to RESIZING to ACTIVE again typically takes a few minutes for the Hetzner migration, plus the time for the health check to confirm VICIdial is responding. It is not a full reinstall, so your data, campaigns, Lead lists, and Call recording files are all preserved on the disk.

What changes and what does not

A resize changes CPU count and RAM on the server. The public IP address stays the same, so your Branded subdomain DNS record does not need to be updated. The TLS certificate stays valid. Your carrier SIP trunk registration comes back automatically after reboot because VICIdial keepalive handles re-registration.

What does not change: your VICIdial data, recordings, agent accounts, campaign settings, and dialing lists. These all live on the disk volume that travels with the server through the resize.

To see which plan tier maps to which box class, visit the pricing page. For the full list of managed features that come with a VICIfast server, read what VICIfast adds to VICIdial. If you also need to add seats alongside the resize, see adding agent seats instantly.