Skip to content
Latchkey

Azure Pipelines "All potential agents are in use" Wait

The job is queued because every matching agent in the pool is already running another job, so there is no free capacity.

What this error means

The job sits in "Waiting for an agent" with a note that all potential agents are in use, until one frees up or times out.

azure-pipelines
The agent request is not running because all potential agents are in use.
Waiting for an agent to become available.

Common causes

Pool saturated

More queued jobs than agents, so jobs wait for a free slot.

Stuck jobs holding agents

Long-running or hung jobs keep agents busy.

Too few agents for peak load

The pool is undersized for concurrent demand.

How to fix it

Add capacity or reduce contention

  1. Add agents to the pool or enable autoscaling.
  2. Cancel hung jobs that are holding agents.

Limit concurrency

  1. Use batching or stage dependencies to smooth peak demand.

How to prevent it

  • If you run on GitHub Actions, self-healing managed runners such as Latchkey auto-retry transient failures and scale capacity, so jobs do not stall waiting for a free agent. On Azure DevOps, size pools for peak load.

Related guides

Tired of flaky CI? Latchkey auto-heals failed jobs and retries them for you. Start free →