Environments

Development, staging and production have different purposes. Development receives new work, staging checks a selected release and production serves users. Always check the environment before an operational action.

Before you start — Identify whether the action concerns development, staging or production. Test availability and the version to test are separate decisions. Read any power action already underway before submitting another request.

Check an environment's availability

Read the installation and operating state before opening the environment’s service links.

1. Open Infrastructure.
2. Choose the development, staging or production view.
3. Read the current state.
4. Open Service to find the application and service links once the environment is available.

An installation not yet provisioned cannot be treated as a paused existing environment. Read the setup state when no resource is available.

Pause a test environment

Pause the selected test environment when it is no longer needed and follow the actual result.

1. Select development or staging in Infrastructure.
2. Open the offered Suspend action.
3. Check Environment and its current state on the confirmation screen.
4. Confirm with Suspend.
5. Return to the overview and follow the state until the request is applied.

The request and the actual stopped state are different stages. Review recorded hosting usage to understand the billed period.

Note — Only development and staging can be suspended; production is never suspended from the workspace. Development and staging can also be suspended automatically when unused, and opening your workspace then starts development again on its own.

Resume a test environment

Resume the test environment before verifying its application, allowing the services to finish starting.

1. Select the paused development or staging environment.
2. Open its start or activation action.
3. Confirm the selected environment with Start.
4. Follow the environment state until the services are available.
5. Open the application and check the task you wanted to test.

The service is reachable and the intended task can be tested in the selected environment.

Resuming restores availability; it does not select a different application version.

Restart an environment

Restart an environment when its servers need to reboot. Every server shuts down and starts again; data, configuration and billing are unchanged.

1. Open the servers view in Infrastructure and select the environment.
2. Click Restart.
3. Check Environment and Status, then confirm with Restart.
4. Wait a few minutes: the interfaces are unreachable until the servers are back up.

Note — Restart is available on production too. A suspended environment must be started before it can be restarted.

Handle a pending action

Follow the current action and its displayed retry condition instead of submitting repeated power changes.

1. Read the message shown when a power change is already underway.
2. Allow that request to finish before submitting another.
3. Follow the retry time if the screen reports a cooldown.
4. Reopen the overview to read the actual state.

The actual environment state is checked after the pending request completes.

Note — Power actions have a five-minute cooldown. Use the time shown by the screen rather than repeatedly submitting the action.

Keep availability separate from publication

Read the version after availability returns so the test concerns the intended release.

1. Confirm that the test environment is running.
2. Read its current version in Deployment.
3. Publish the intended version to staging when needed.
4. Test the user journey before advancing that version to production.

Opening or restarting an environment does not by itself publish the latest project changes.

When staging starts, it can serve the version that was already published there. Opening the application proves availability; reading Deployment identifies its version. If another version must be tested, use the publication workflow after startup rather than treating the start action as a release.

Resolve an availability problem