Error codes reference

This page lists every error code of a run, grouped by area, with what it means and what to do. Use it to troubleshoot a run that failed or did not start.

This page is generated from the product code; do not edit by hand.

How to read an error

When a run or an action fails, the Station and the Console show a short message and an explanation. Under Show details they show the exact text, which starts with a code, for example missing-variables: …. Find the code below.

• You see is the message on screen.
• What to do lists the fixes, the most likely first.
• Running again unchanged can work says whether a second try, without changing anything, can succeed.
• Shown as is how the message is marked: Error, Warning or Information.

Starting a run

environment-unreachable

You see: “The environment did not answer”

Before a run, the Station calls the environment’s health address. It did not answer from this PC, so the run stopped before any test started.

What to do:

1. Check that this PC is on the network or VPN that reaches the environment.
2. Open the environment in a browser to see whether it is up.
3. Run again once it answers.

Running again unchanged can work: yes. Shown as: Error.

missing-variables

You see: “Variables are missing”

The workload needs variables or secrets that are not set for this environment. The details list their names; values are never shown.

What to do:

1. Ask a project admin to set them in the Console, in the environment’s variables or secrets.
2. Run again once they are set.

Running again unchanged can work: no. Shown as: Error.

not-configured

You see: “This workload or environment is not set up here”

The Station does not know this workload or environment, or the workload is not enabled on that environment.

What to do:

1. Run again: the Station fetches the latest configuration from the Console when a run starts.
2. Ask a project admin to enable the workload on the environment.

Running again unchanged can work: no. Shown as: Error.

unknown-profile

You see: “This profile does not exist”

The chosen profile is not in the workload’s manifest on this branch. The details list the profiles it has.

What to do:

1. Pick one of the listed profiles and run again.

Running again unchanged can work: no. Shown as: Error.

override-not-allowed

You see: “This variable cannot be overridden”

Secrets and variables starting with ORBIT_ cannot be changed for a single run. The details name the variable.

What to do:

1. Remove it from the overrides and run again.

Running again unchanged can work: no. Shown as: Error.

archived

You see: “Archived in the Console”

The project, workload or environment of this run was archived in the Console, so no new runs start on it. Its past runs stay visible.

What to do:

1. Ask a project admin to restore it in the Console if it is still in use.
2. Otherwise pick another workload or environment.

Running again unchanged can work: no. Shown as: Error.

rerun-not-allowed

You see: “This re-run is not allowed”

A re-run follows the same rules as a new run. The chosen environment is not a target of this workload, or you may not run on it.

What to do:

1. Pick an environment the workload runs on and you can run on.
2. Ask a project admin for access if you need that environment.

Running again unchanged can work: no. Shown as: Warning.

disk-low

You see: “The disk is almost full”

The drive with the Station’s data folder has less free space than the Station keeps as a minimum, so the run did not start. A run that fills the disk would fail half-way.

What to do:

1. Open Settings, Disk, and remove caches or old run files you no longer need.
2. Free space on the drive, or lower the minimum in Settings, Disk.
3. Run again.

Running again unchanged can work: yes. Shown as: Error.

Fetching and installing the test code

checkout-failed

You see: “The test code could not be fetched”

Git could not fetch or check out the workload’s repository.

What to do:

1. Check the network connection to GitHub.
2. Check that the branch, tag or commit exists.
3. If it keeps failing, copy the details and send them to a workload maintainer.

Running again unchanged can work: yes. Shown as: Error.

manifest-invalid

You see: “The workload’s manifest has errors”

The orbit.workload.yaml file on this branch does not match what the Station expects. The details list each problem.

What to do:

1. Ask a workload maintainer to fix the manifest on this branch.
2. Or run a branch whose manifest is valid.

Running again unchanged can work: no. Shown as: Error.

install-failed

You see: “Installing the workload failed”

Installing the workload’s dependencies or Playwright browsers failed. The details name the step; the run log has the full output.

What to do:

1. Open the run log for the package manager’s message.
2. If a download failed, check the network or the proxy and run again.
3. If it says the Station could not read a repository, run Diagnostics, then ask a project admin to check that the workload’s repositories are in the Orbit organisation.
4. If the lockfile is missing or out of date, ask a workload maintainer to fix the repository.

Running again unchanged can work: yes. Shown as: Error.

prepare-timeout

You see: “Preparing took too long”

Fetching the code and installing dependencies and browsers did not finish in the time allowed. Time spent waiting for a lock does not count.

What to do:

1. Run again: a second run reuses what was already downloaded.
2. If it happens again, check the network speed or the proxy.

Running again unchanged can work: yes. Shown as: Error.

While the tests run

run-timeout

You see: “The tests ran out of time”

Playwright ran longer than the profile allows, so the Station stopped it. Results of the tests that finished are kept; the ones that did not finish are marked Not run.

What to do:

1. Open the report to see which tests were slow or stuck.
2. Run fewer tests, or ask a workload maintainer to raise the profile’s timeout.

Running again unchanged can work: yes. Shown as: Error.

collect-timeout

You see: “Reading the results took too long”

Playwright finished, but reading its results took too long. Results sent during the run are kept; some may be missing.

What to do:

1. Open the HTML report for the complete results.
2. Run again if results are missing.

Running again unchanged can work: yes. Shown as: Warning.

station-restart

You see: “The Station stopped during the run”

The Station closed, or the PC restarted, while this run was going. Results of the tests that finished are kept.

What to do:

1. Run again.
2. If most tests passed, re-run only the failed ones.

Running again unchanged can work: yes. Shown as: Warning.

Locks and test accounts

lock-held

You see: “Another run holds a lock”

A selected test changes a shared part of the environment, and another run is using it. This run waits until the lock is free.

What to do:

1. Wait, or cancel and run tests that need no lock.
2. Ask the person who holds it how long they need.

Running again unchanged can work: yes. Shown as: Information.

lock-lost

You see: “The run lost its lock”

The run held a lock on a shared part of the environment and the Console stopped confirming it. Another run may now change the same data, so this run stopped.

What to do:

1. Check the network connection to the Console.
2. Run again.

Running again unchanged can work: yes. Shown as: Error.

account-unavailable

You see: “No test account is free”

The workload needs accounts of a role from the test account pool, and other runs hold every one of them. The run waits until one is free; when the pool has fewer active accounts than the run needs, it stops instead.

What to do:

1. Wait, or cancel and run later.
2. Ask a QA lead who holds the accounts, or to release a lease that is stuck.
3. Ask a project admin to add accounts of this role in Project settings > Test accounts.

Running again unchanged can work: yes. Shown as: Information.

account-lease-lost

You see: “The run lost its test account”

The run leased an account from the pool and the Console stopped confirming the lease, or a QA lead released it. Another run may now sign in with the same account, so this run stopped.

What to do:

1. Check the network connection to the Console.
2. Run again.

Running again unchanged can work: yes. Shown as: Error.

station-too-old

You see: “This Station is too old for this workload”

The workload takes its sign-in accounts from the test account pool on this environment, which needs a newer Station. The run did not start, so no shared account was used.

What to do:

1. Update the Station: Settings → About shows the update when one is ready.
2. Run again once it is updated.

Running again unchanged can work: no. Shown as: Error.

Console, sign-in and access

console-unreachable

You see: “The Console could not be reached”

The Station could not talk to the Console. A workload that already ran on this Station can still run offline.

What to do:

1. Check the network connection.
2. Run again: when the Console is still unreachable, the run starts offline.

Running again unchanged can work: yes. Shown as: Error.

console-refused

You see: “The Console refused the run”

The Console did not accept this run, for example because the workload, the environment or your role changed.

What to do:

1. Run again: the Station fetches the latest configuration from the Console when a run starts.
2. If it is refused again, copy the details and send them to a project admin.

Running again unchanged can work: no. Shown as: Error.

not-allowed

You see: “You may not run here”

Your role does not allow this action on this environment, or your access was removed.

What to do:

1. Ask a project admin for a role that can run on this environment.

Running again unchanged can work: no. Shown as: Error.

signed-out

You see: “You are signed out”

Nobody is signed in on this Station, or your session ended. Runs use your Console account.

What to do:

1. Sign in again, then run.

Running again unchanged can work: no. Shown as: Warning.

Working offline

source-unavailable-offline

You see: “The test code is not on this PC”

The Station is offline and has never fetched this workload, or this branch, so it has no copy of the tests to run.

What to do:

1. Connect to the network and run once online.
2. Or pick a branch that has already run on this Station.

Running again unchanged can work: no. Shown as: Error.

config-unavailable

You see: “The Station has no configuration”

The Station could not get its configuration from the Console and has no saved copy for you.

What to do:

1. Check the network connection.
2. Connect once online before running offline.

Running again unchanged can work: yes. Shown as: Error.

offline-grace-ended

You see: “Offline time is up”

This environment allows a limited number of days offline since the Console last confirmed your access, and that time has passed.

What to do:

1. Connect to the network so the Console can confirm your access, then run again.

Running again unchanged can work: no. Shown as: Error.

secrets-unavailable-offline

You see: “The secrets are not on this PC”

The run needs secrets, and this Station has none saved for this workload and environment. They are saved only after an online run.

What to do:

1. Connect and run once online; later runs can work offline.

Running again unchanged can work: no. Shown as: Error.

Choosing tests and the test list

no-catalog

You see: “No test list yet”

Choosing single tests needs the workload’s test list, which the Station builds on the workload’s first run.

What to do:

1. Run the whole workload once, then choose tests.

Running again unchanged can work: no. Shown as: Information.

unknown-test

You see: “A chosen test is no longer in the list”

A selected test is not in the workload’s test list any more, for example because it was renamed or removed.

What to do:

1. Open the Explorer again and choose the tests again.

Running again unchanged can work: no. Shown as: Error.

selection-too-large

You see: “Too many tests chosen one by one”

The chosen tests make a test filter longer than Windows allows on one command line, so the run did not start.

What to do:

1. Choose fewer tests, or whole files and describes, which take less room.
2. Run a profile that covers these tests instead.

Running again unchanged can work: no. Shown as: Error.

refresh-running

You see: “The test list is already being refreshed”

Another Refresh from main of this workload is still running on this Station. Two refreshes at once would only repeat the same work.

What to do:

1. Wait for the running refresh to end; the Explorer then shows the new list.

Running again unchanged can work: yes. Shown as: Information.

catalog-refresh-failed

You see: “The test list could not be refreshed”

The Station could not fetch the latest code or list its tests, so the Explorer keeps the earlier list. A workload whose Playwright config needs a secret to load cannot be listed without it, because a refresh never uses secret values.

What to do:

1. Read the details for the step that failed.
2. Try again; a second attempt reuses what was already downloaded.
3. Or start a run on the default branch, which also updates the list.

Running again unchanged can work: yes. Shown as: Error.

catalog-refresh-timeout

You see: “Refreshing the test list took too long”

Fetching the code, installing dependencies and listing the tests did not finish within 10 minutes, so the Explorer keeps the earlier list.

What to do:

1. Try again: a second attempt reuses what was already downloaded.
2. If it happens again, check the network speed or the proxy.

Running again unchanged can work: yes. Shown as: Error.

Station files, history and support

history-unavailable

You see: “No history for this test”

The Station could not read the history of this test. Its own results show when it has some; the Console part needs a connection and a Console that knows the test.

What to do:

1. Connect to the network and open the history again.
2. Run the test once.

Running again unchanged can work: yes. Shown as: Information.

bundle-failed

You see: “The support bundle was not made”

The Station could not build the support bundle. Nothing was sent anywhere: the bundle stays on this PC.

What to do:

1. Try again.
2. Use “Copy summary” on the Diagnostics page and paste it to the person helping you.

Running again unchanged can work: yes. Shown as: Error.

not-synced

You see: “Not on the Console yet”

Some of this run’s results, report or log still wait to be sent to the Console. Removing its files now would lose them.

What to do:

1. Wait until Sync shows everything sent, then try again.
2. Open Sync to retry items that failed to send.

Running again unchanged can work: yes. Shown as: Warning.

in-use

You see: “In use right now”

A run is using these files at the moment, so the Station leaves them alone. Nothing was removed.

What to do:

1. Wait until the run finishes, then try again.

Running again unchanged can work: yes. Shown as: Warning.

not-found

You see: “Not found on this Station”

What you asked for is not here any more. It may have been deleted, from this page in another tab or by the cleanup.

What to do:

1. Reload the page to see what is there now.

Running again unchanged can work: no. Shown as: Information.

Other errors

An error without a code in this list is shown as “Something went wrong”. This error has no explanation yet. The details hold the exact message.

What to do:

1. Copy the details and send them to a workload maintainer.

Failure categories of a test

A failed test is put in one category from its error message. The category is a first hint of where to look.

Category	What it usually means
Assertion	The page did not show what the test expected: compare the expected and received values.
Element not found	An element was missing or matched more than one: check whether the screen changed or the locator is too broad.
Timeout	Something took longer than allowed: the system may be slow, or the test waited for something that never appeared.
Navigation	The page navigated or reloaded while the test used it: wait for the navigation to finish first.
Network	A request did not reach the server: check that the environment is up and reachable from this PC.
Sign-in	The application refused the login or the session: check the test account and whether its password changed.
Environment	The application answered with a server error: the environment may be down or under maintenance.
Test code	The test code itself threw an error, such as a typo, a missing value or a changed helper: read the stack.
Other	No known pattern matched: read the error message and open the trace.