When servicing an interrupt, complete it before servicing another
If a developer is already working in "interrupt mode" (see InterruptsUnjamBlocking) on a critical issue, don't put that work aside until it is complete or until that issue itself becomes hopelessly tangled.
When a role is understaffed or under localized schedule pressure, it could get “interrupt” calls to prioritize resources it provides from several roles in a short time space. Stopping work on one priority call to service another leads to thrashing and slows the overall rate of work. Do not stop work on a current priority call until it is complete or no further work can be done on it. Note that if we ever were to achieve the ‘ideal’ situation discussed in InterruptsUnjamBlocking, we would be just a hair’s breadth away from this situation.