On Wed, Nov 9, 2016 at 4:54 PM, Peter Geoghegan <p...@heroku.com> wrote: > It's more complicated than that. As I said, I think that Knuth > basically had it right with his sweet spot of 7. I think that commit > df700e6b40195d28dc764e0c694ac8cef90d4638 was effective in large part > because a one-pass merge avoided certain overheads not inherent to > polyphase merge, like all that memory accounting stuff, extra palloc() > traffic, etc. The expanded use of per tape buffering we have even in > multi-pass cases likely makes that much less true for us these days.
Also, logtape.c fragmentation made multiple merge pass cases experience increased random I/O in a way that was only an accident of our implementation. We've fixed that now, but that problem must have added further cost that df700e6b40195d28dc764e0c694ac8cef90d4638 *masked* when it was commited in 2006. (I do think that the problem with the merge heap maintenance fixed recently in 24598337c8d214ba8dcf354130b72c49636bba69 was the biggest problem that the 2006 work masked, though). -- Peter Geoghegan -- Sent via pgsql-hackers mailing list (email@example.com) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers