Markus Wanner <markus@bluegap.ch> wrote:
> I don't remember reading about predicate locking in the paper I
> read. Either he didn't cover that in his first implementation
> (based on page level locking), or I've simply re-used that part of
> my in-brain-memory.
If you read the first paper but not the second, all you would have
seen was the mention that predicate locking was covered because
Berkeley DB uses page level locks for everything. The second paper
gets into a little more detail on the subject, with references to a
good overview of the topic as well as some authoritative papers.
-Kevin