Long running locks could be problematic

[rsyring]

PostgreSQL uses one process per connection, so the number of connections doesn't scale that well. It's possible that if you needed to do a lot of locks across a lot of processes that this might become a problem.

Some more reading on this here: https://news.ycombinator.com/item?id=21536698

Not really sure what, if anything, PALs should do about this, maybe just some documentation around the problem would be helpful.