[bobbojones@test parking] > select count(1) from mbta_passes; +----------+ | count(1) | +----------+ | 179399 | +----------+ 1 row in set (27.65 sec)
That’s just not cool. It took 27.65 seconds to count the number of rows? Using a small subset of test data? On a test server that nobody else was using at the time of that query? That is just not cool.
Awesome. I’m working on SQL IO performance issues, right now, with a client. I can think of quite a few reasons it would take so long to run that query. SQL Tuning is Fun!
How many of those reasons apply under the circumstances I described, though? Yes, it’s an InnoDB table, and our test servers are not necessarily the powerhouses our production servers are, but that’s still a very small number of rows to have counted in that period of time, neh?
this proves public transportation is too slow