I should add to comment #11, this obviously isn't constant or it would likely be a single local bottleneck. As near as I can tell its nearly random, but higher usage times do make it more likely. And, of course, if I try to run the very same query that took 4+ minutes to run and was killed, this time it will return in a reasonable number of milliseconds.
I should add to comment #11, this obviously isn't constant or it would likely be a single local bottleneck. As near as I can tell its nearly random, but higher usage times do make it more likely. And, of course, if I try to run the very same query that took 4+ minutes to run and was killed, this time it will return in a reasonable number of milliseconds.