Re: is it possible to make this faster?
От | Merlin Moncure |
---|---|
Тема | Re: is it possible to make this faster? |
Дата | |
Msg-id | b42b73150605261240k64f3cf3ej5c2feebbe9c67c89@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: is it possible to make this faster? (Tom Lane <tgl@sss.pgh.pa.us>) |
Ответы |
Re: is it possible to make this faster?
|
Список | pgsql-performance |
On 5/26/06, Tom Lane <tgl@sss.pgh.pa.us> wrote: > > mysql> select user_id, acc_id, max(sample_date) from usage_samples group by 1,2 > > 939 rows in set (0.07 sec) > > 0.07 seconds is not impossibly out of line with my result of 0.15 sec, > maybe your machine is just 2X faster than mine. This is a 2.8GHz dual > Xeon EM64T, what are you testing? You said "less than 20 msec" before, > what was that on? 600 mhz p3: 70 ms, 1100 ms slow case 1600 mhz p4: 10-30ms (mysql timer not very precise) 710ms slow case quad opteron 865: 0 :-) dual p3 1133 Mhz xeon, mysql 4.0.16: 500 ms using steinar's 'substitute group by' for pg I get 40ms on the p3 and low times on all else. your time of 150 ms is looking like the slow case on my results. merlin
В списке pgsql-performance по дате отправления: