compartment-disposed: Disposed isolation instance

Cause

Code attempted to use a Compartment or ESM execution instance after dispose(), or disposal canceled unfinished work. Common cases:

• dispose() canceled an unfinished classic script evaluation.
• The ESM entry module was still loading during dispose(), so its entry promise was rejected.
• After the sandbox was disposed, code still used the resource transformer or fetch it had retained to insert scripts or styles or send requests, such as a callback that fires after the micro app unmounts.
• mount() or nodeTransformer was called on a sandbox controller after its dispose(), or mount() was still running when dispose() happened.

Troubleshooting

1. Use the stack trace to identify whether script evaluation, module import, a dynamically inserted resource, or another instance operation failed.
2. Check the order of dispose() and asynchronous work. Look for timers, callbacks, or pending promises that still use the old instance.
3. Check whether the remount flow tries to reuse an instance that has already been disposed.

Solution

If the task needs to finish, wait for evaluation or import to complete before disposing the instance. If disposal is an intentional cancellation, handle the rejected promise and stop subsequent work.

dispose() is terminal cleanup. Unlike unmount(), it does not allow remounting the same instance. Create a new instance and mount it when the micro app needs to run again.

Related

• Error codes and solutions
• compartment-disposed：隔离实例已销毁