Skip to content

Priority Agent

When to Use a Priority Agent

Priority Agents are recommended in scenarios such as:

  • Collaboration use case.
  • Work-from-home or hybrid work environments.
  • File caching jobs.

They ensure that files are reliably uploaded and distributed across the system without risking data loss or bandwidth bottlenecks.

How Priority Agents affect jobs

  • File upload prioritization. New or updated files are uploaded to the Priority Agent first, before being seeded to other peers utilizing p2p. This guarantees that the Priority Agent always has a complete copy.
  • Prevents file dehydration. Files in cache, on end-user folders or in TSS folders won't dehydrate until the Priority Agent has received all file pieces. This ensures at least one reliable copy exists in the job.
  • Prevents premature Agent shutdown. If a user stops their Agent while some files are being uploaded to Priority Agent, the system ensures they reach the Priority Agent first.

Enable Priority Agent

When creating or editing a job, in the AGENTS tab, enable Priority Agent option for select Agents.

General requirements

  • Must not be a selective sync Agent.
  • Must remain always online.
  • Ideally, placed on a fast and reliable network.
  • High Availability or Scaleout groups can be assigned as Priority Agents.
  • In hybrid/cache jobs, the Primary Storage is selected as Priority Agent by default (this can be unchecked).
  • Multiple Priority Agents can be selected for redundancy. In this case file upload prioritization is targeted at at least one of them before file can be dehydrated or the Agent can be disconnected.

If an Agent cannot reach the Priority Agent, it will report the error.

For version older than 4.0, in the Agent profile, using custom parameter priority_peers, with AgentID as value. Use semicolon as separator.

What Happens Without a Priority Agent If no Priority Agent is configured:

Files may dehydrate before uploading, leaving no reliable copy. No error or warning is reported.

Bandwidth bottlenecks can occur when other Agents in the job pull data from a single user. As a result, files may take significantly longer to appear on other Agents in the job, as well as on the central peer (e.g., SMB access to primary storage).

If that user disconnects before at least one Agent gets the full file copy, files may become unavailable.