Skip to main content
This reference explains how Nonce Agent discovers miners and how miner data reaches Nonce.

​
Overview

At a high level, the flow looks like this:
  1. You create a scan task from the Nonce UI.
  2. The Agent polls the server for tasks and starts the scan.
  3. The Agent probes the configured IP range and tries to identify miners one IP at a time.
  4. When a miner is found, the Agent immediately fetches miner details and sends a realtime update.
  5. The backend ingests the uploaded raw miner data and updates the miner records shown in Nonce.
  6. The Agent continues running background refresh jobs so miner data stays up to date after the scan.

​
Scan Behavior

​
Manual scan

  • Manual scan is designed to provide feedback quickly.
  • In current product behavior, some miners may appear before the full scan task finishes.
  • When miners are discovered, the Agent also adds them into its ongoing monitoring list.

​
Automatic scan

  • Saved auto-scan IP ranges are rescanned on a schedule.
  • This allows newly connected miners to be discovered later even if you do not trigger another manual scan right away.

​
How Long Until Miner Data Appears?

The exact time depends on IP range size, network quality, miner responsiveness, and whether your deployment changed the default Agent intervals. In a typical setup, these are good expectations:
  • Agent installation confirmation: usually within a few seconds after you run the install command and click confirm
  • Manual scan task pickup: usually within about 5 to 10 seconds
  • First miner data after a manual scan starts: often within tens of seconds, because discovered miners are uploaded through the realtime path
  • Regular full miner refresh: about every 5 minutes by default
  • Automatic rescan of saved IP ranges: hourly by default
Task success and refreshed miner rows in the UI are not always the same moment.

​
What Affects Scan Duration?

These factors have the biggest impact:
  • IP range size: scanning /24 is much faster than scanning several large ranges at once
  • Miner response speed: slow or overloaded miners take longer to identify
  • Network quality: routing issues, packet loss, or firewall rules can slow detection
  • Miner credentials: incorrect web passwords can prevent metadata collection for some miner models
  • Retries: the Agent retries failed miner data fetches, which improves success rate but can increase total wait time