Hi all,
Just bought a new server, which was supposed to improve the performance of our app.
The new machine has the same OS (Redhat 8), same MySQL (4.0.18), and same my.cnf.
The problem is that this (frequently run) query, actually runs 41% slower!
select * from LEG L, LEG_DETAIL D, DEAL_LEGS G
where L.latest_id = D.latest_id and D.latest_id = G.latest_id
group by L.dealid limit 750000;
If I remove the "GROUP BY", then the new box returns the results 14% faster than the
old server.
(So what's so special about "GROUP BY"?)
I ran the benchmark scripts on both boxes, and it seems the following operations are
slower on the new server, for some reason?
count_distinct (1000)
count_distinct_2 (1000)
count_distinct_group (1000)
count_distinct_group_on_key (1000)
count_distinct_group_on_key_parts (1
count_distinct_key_prefix (1000)
count_group_on_key_parts (1000)
count_on_key (50100)
select_distinct (800)
select_group (2911)
Can anyone suggest why the new machine might be slower, ONLY IN THE ABOVE areas?
Just in case these numbers help, here are lines from the RUN file, for two of the slow
operations above....
Operation seconds usr sys cpu tests
count_distinct_group 19.00 1.17 0.08 1.25 1000
count_distinct_group (new box) 26.00 0.39 0.26 0.65 1000
count_distinct 12.00 0.51 0.02 0.53 1000
count_distinct (new box) 15.00 0.10 0.01 0.11 1000
Any advice at all, would be very much appreciated!
Thanks,
Tony
***********************************************************************
Bear Stearns is not responsible for any recommendation, solicitation,
offer or agreement or any information about any transaction, customer
account or account activity contained in this communication.
***********************************************************************
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
