DB corruption within minutes

[Matthias Andree]

David Relson <relson at osagesoftware.com> writes:

> What's the real problem here?  That BerkeleyDB doesn't have real locks
> unless transactions are used?  That Linux has a lock problem?  Something
> else?

I haven't tried mandatory locking on Linux, but advisory locking works
exactly as documented, no nasty surprises in /this/ area in four years.

BerkeleyDB seems to need transactions to be able to recover from certain
error conditions.

-- 
Matthias Andree