journal. If I could flush the dirty pages first, then the actual commit
is short enough to be effectively instantaneous and there's little
chance of an interrupt arriving at the same time.
Would it be possible to expose this first phase via a C or SQL API?
Thanks!
--
Michael Tharp
rPath, Inc
Hullo,
I have a package manager that consists of a sqlite database combined
with a filesystem journal that is created during system updates. The
journal contains enough information to clean up the filesystem if an
update is interrupted by e.g. a power failure. Naturally, these two
journals
On 02/14/2010 01:09 AM, Simon Slavin wrote:
read the distinctions between the different types of 'BEGIN'.
I don't see any mention of flushing behaviour, just earlier locks. I
guess what I'm asking for is a lightweight two-phase commit, which seems
easy enough to do given the current