State consistency after workflow abort

[lmagyar]

Currently the workflow is not reloaded until the next incoming workflow related request or reactivation reminder (2 minutes), until this the non workflow related requests will read an obsolete state, that will be reloaded later.
Should we reload State after abort immediately (but not run, that can cause abort again), or let non workflow related requests to prepare the workflow state and implicitly the grain state?

[lmagyar]

This should be treated like InconsistentStateException, ie. deactivate. This means ReadState on the callback interface is unnecessary.