Re: Which server hardware would you pick?
Rick, thanks for the feedback. The 2nd server does look like it should perform better - much better. But it doesn't.
The two servers are actually setup in a multi-master configuration, so updates can be made on either.... so the data is exactly the same on both.
Server #2, which should be faster goes into a type of backlog state as it can't keep up with the volume of activity (which in my opinion is very low). Normal connections may be 70, but it will backlog and there will be 800+ connections and it is like there is a lock preventing anything from moving.
I have a couple of Innodb tables with a couple million records, which I purge after 7 days.
You can see the performance differences clearly in a script runs and purges 500 records at a time from each table.
Server #1 time (average over several runs)
500 records removed from tblquarantine in 0.017232 seconds.
500 records removed from tblmsgs in 0.805116 seconds.
Server #2 time (average over several runs)
500 records removed from tblquarantine in 1.452923 seconds.
500 records removed from tblmsgs in 1.288616 seconds.
Server #2 shoud be faster, but everything shows it is slower - a lot slower - and can't keep up under the same load as Server #1.
If anyone has some tips I can try to determine where the bottleneck is I would greatly appreciate it.
Subject
Views
Written By
Posted
1632
November 14, 2014 08:24PM
860
November 15, 2014 08:46PM
Re: Which server hardware would you pick?
787
November 16, 2014 06:25PM
765
November 17, 2014 03:42PM
967
November 18, 2014 09:08AM
735
November 19, 2014 05:02PM
Sorry, you can't reply to this topic. It has been closed.
Content reproduced on this site is the property of the respective copyright holders.
It is not reviewed in advance by Oracle and does not necessarily represent the opinion
of Oracle or any other party.