How to turn on System Performance logging in VICIdial

VICIdial does not collect server performance data by default — you have to explicitly enable it per server, and until you do the Server Performance Report will show nothing, no matter how long you wait.

Where to turn it on

The setting lives in the server record itself, not in a global configuration screen. Navigate to the Administration section and open the record for the server you want to monitor. Find the field labeled System Performance and change its value to Y. Save the record. That is the entire configuration step — there is no service restart required and no cron job to schedule.

Once the flag is set to Y, VICIdial begins sampling the box every five seconds and writing the results to its logging tables. The first data point appears almost immediately. The Server Performance Report then becomes usable as soon as you have a meaningful window of data — typically after a full shift has passed.

The setting is per server, not global

If you run more than one Asterisk server under the same VICIdial installation, you need to enable System Performance on each server record individually. A flag set on server A does not propagate to server B. In multi-server setups, go through each server record and confirm the setting is Y on all boxes you want to track. This matters particularly for operations that have split campaigns across servers: without per-server enablement you will have blind spots in your monitoring.

The server-specific reports table in the Administration section also surfaces useful at-a-glance metrics per server — load average, CPU load percentage, active Asterisk Channel count, and the highest hard-drive use percentage. These columns rely on the same underlying logging that the System Performance flag enables, so they too will stay blank on any server where the flag is off.

sequenceDiagram
  participant Admin as VICIdial Admin UI
  participant DB as Server Record in DB
  participant Logger as VICIdial Logger Process
  participant Report as Server Performance Report
  Admin->>DB: Set System Performance = Y
  DB-->>Logger: Flag polled on next cycle
  Logger->>Logger: Sample every 5 seconds
  Logger->>DB: Write CPU, load, channel data
  Admin->>Report: Open report for date range
  Report->>DB: Query logged samples
  DB-->>Report: Return data points
  Report-->>Admin: Display graphs and summary

Why the report cannot be downloaded

The Server Performance Report is one of the few reports in VICIdial that has no download or export option. If you want to keep a record of a particular shift's graphs and summary numbers, take a screenshot before the data ages out of your retention window. This is worth building into your shift-close checklist if you are tracking performance trends over weeks rather than days.

The lack of an export is a deliberate constraint of the report, not a bug. Because the data is sampled every five seconds across a full shift, the table that backs it can grow large. The report is built to be viewed interactively in the browser for a chosen date range rather than pulled into a spreadsheet. If you need long-term trend data, your best option is to build a habit of weekly screenshots at a consistent time — same shift, same day of week — so you have a visual baseline to compare against.

What you get once it is on

With logging active, the report gives you a shift summary covering total calls handled, off-hook minutes, average and peak system load, average Concurrent calls via the Asterisk channel count, and the USER/SYSTEM/IDLE CPU split. Below the summary, a graph plots all four key metrics — processes, load, channels, and CPU — over the date range you select using the YYYY-MM-DD HH:MM:SS format.

That five-second sampling interval is fast enough to catch brief spikes that a one-minute average would completely smooth away. A keepalive loop that hammers the CPU for thirty seconds and then calms down would be invisible in most monitoring tools but shows up clearly in the VICIdial performance graph as a transient spike in the load and SYSTEM lines. The granularity is one of the reasons this report is worth enabling even if you only look at it reactively when something feels wrong.

The guide to monitoring VICIdial server health and capacity covers what to do with the data once it starts arriving — what thresholds to watch and how often to check.

For a tour of what the graphs show and how to read the four series in context, see how to read the Server Performance Report graphs.

On a managed VICIfast server, performance logging is turned on from the moment the box provisions — you do not need to find the setting and flip it yourself. Start a VICIfast trial and your server is live and logging in under 40 seconds.