Antony Blakey wrote: > -- snip -- > > I think that keeping these two points explicit results in simpler and > more transparent reasoning around issues *such as* atomicity - in > particular, why it will never happen. > > As far as rollback is concerned, the problem is that concurrent > accessors can see the intermediate state. There is no way to fake ACID > without serializing all access (not always feasible), and even then > there are failure modes in a clustered situation equivalent to > transaction monitor lockup. >
For my particular purpose, where I will have a background check for conflicts / referential integrity, I don't need atomicity. Hence an attempt at rollback in some fashion would be enough; if it fails occasionally because of concurrent access, I'll deal with it. Tim
