Hosted Pipelines times out before the first command starts (agent-service.queue.timeout)

[Nodar Shurghaia]

Hello,

We are setting up a new Bitbucket Cloud repository and a very small pull-request Pipeline cannot start.

Four runs have failed before the first script command executes. Each waited roughly 45 minutes, recorded zero build seconds/minutes, then failed with:

- Error key: agent-service.queue.timeout

- Message: Could not process step progress updated event.

The Pipeline uses atlassian/default-image:4 and has one step that only checks for five tracked files. It does not install dependencies, build Docker images, deploy, use secrets, call external services, or run Terraform. Git access, repository access, Pipelines enablement, the pull-request selector, and the YAML have been checked. The exact checks pass locally.

Could this indicate a hosted-runner allocation or new-workspace entitlement problem? Is there an admin-side diagnostic or setting we should check? Also, would upgrading from Free to Standard change hosted-runner allocation, or only support and build-minute limits?

Thank you.

[Arkadiusz Wroblewski]

Hi @Nodar Shurghaia , welcome to the Community!

This isn't a script issue, `agent-service.queue.timeout` means Bitbucket failed to provision a runner, timing out after 45 minutes before your step ever started.

Check that your workspace still has available build minutes. Then, try running a minimal custom pipeline with just `echo "test"` to see if it starts.

If that test fails as well, let me know here so I can flag this post for the Atlassian team to take a look!

Best,

Arek 🤠