15.2. A model of a crash
A simple model of a crash will be taken initially, a fail–stop model. We shall assume that a crash happens at some instant rather than over a period of time. It results in the instantaneous loss of volatile state: the processor registers, the cache and memory management unit, and the (volatile) main memory. Any changes that have been made to persistent state, for example in any non-volatile memory and disk-based storage, are assumed to be correct, but may be incomplete. By making this simplifying assumption we avoid considering how incorrect system software may behave before finally causing a crash. In Section 15.7 we take a more realistic look at the likely behaviour of failing software.
Schlichting and Schneider (
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access