On Wed, Jun 27, 2018 at 11:35 AM, Luca Ferrari <fluca1978@gmail.com> wrote:
> If I then disconnect and reconnect I'm able to issue the select and
> get back the results. But if I issue a reindex I got the same error
> and the table "becames unreadable" for the whole session.
> On 10.3 the table is never locked for the session, that is I can
> create the index, I can query the table and get the results, but I
> cannot reindex. However, even after a reindex, it does allow me to
> select data from the table.
>
> So my question is: why this behavior in later PostgreSQL?
It might have something to do with the changes to parallel CREATE
INDEX. It changed how we tracked whether or not an index could be used
because it was currently undergoing reindexing. This is supposed to
make no difference at all, but there was one bug that could cause us
to consider an index irrevocably unusable.
Do you find that the issue goes away if you set
max_parallel_maintenance_workers=0 on v11/master?
--
Peter Geoghegan