On 6 October 2016 at 18:33, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I'm a bit confused about exactly what the context is here. AFAICS,
> the fragment you quoted should work as you expect, as long as the
> table always exists beforehand. Then, the DROPs serialize the
> transactions' access to the table and all is well. On the other hand,
> if the table *doesn't* exist beforehand, there is nothing to serialize
> on and the behavior Adrian exhibited is what I'd expect.
I accept that this is how things are. I'm just surprised that "DROP
TABLE IF EXISTS" doesn't do the exists-test at commit time, rather
than effectively being "DROP TABLE IF
EXISTED-AT-SOME-RANDOM-POINT-IN-THE-PAST".
At the end of the day this isn't a massive deal - I can simply add
exception code around the failure, as Francisco suggested, or add
oplocks around the code (as per Kevin), or (I suppose) I could do an
individual transaction to CREATE TABLE IF NOT EXISTS as a separate
transaction before I start; it's just something that caught me out
because I didn't expect it to be a problem.
Geoff