On Thu, Sep 29, 2011 at 10:22 AM, Robert Haas <robertmhaas@gmail.com> wrote:
> I can't really explain why people seem to keep wanting to create
> hundreds of thousands or even millions of tables, but it's not like
> MauMau's customer is the first one to try to do this, and I'm sure
> they won't be the last. I don't want to de-optimize the more common
> (and sensible) cases too much, but "slow" still trumps "fails
> outright".
Yeah -- maybe baby steps in the right direction would be track cache
memory usage and add instrumentation so the user could get a readout
on usage -- this would also help us diagnose memory issues in the
field. Also, thinking about it more, a DISCARD based cache flush
(DISCARD CACHES TO xyz) wrapping a monolithic LRU sweep could help
users deal with these cases without having to figure out how to make
an implementation that pleases everyone.
merlin