Not sure about the severity of this, as it's a pure API function that we don't expose a UI for.

Calling flag_reset_flag() will leave zombie fields on flagging entities, as the {flagging} table entries are deleted but the entity and field APIs are not invoked.

Comments

joachim’s picture

This should probably use queue API to prevent timeouts: see #89181: Use queue API for node and comment, user, node multiple deletes for reference.