Re: PostgreSQL vs MySQL Throughdown! Teams?

["markw-4" <user=5@nabble.blu.org>]

 Your post sort of states why I think this would be 
[fun|good|useful|educational] 

Everyone has a benchmark and for some reason it never really compares 
apples to apples and oranges to oranges. There is always some aspect of it 
that isn't coherent. 

I think a "real life" benchmark that can be (is) run against MySQL and 
PostgreSQL on the same machines, and not just the huge power servers, but 
the conservative median servers that are used in development and small to 
medium companies would be useful for a lot of people. 

The problem with TPC is that, last I recall, MySQL did not support enough 
of the SQL standard to implement it. 

Lastly there is an amount of real experience that is gained by doing this 
sort of thing on your own. It is one thing to read the results of someone 
else's benchmark, but it is quite another thing to be writing/performing 
the benchmark and seeing the effects on your own and really equating the 
nature of the test and results. Having the chance to tune the queries for 
better results and seeing the effect of different query strategies. 
Professionally, I have not benchmarked MySQL in a few years and I'm going 
to do it for my own edification. It would be cool if more people can get 
involved and make it a broader and more comprehensive test. 


> Message: 5 
> Date: Wed, 16 Jan 2008 10:09:02 -0500 
> From: Seth Gordon <[hidden email]> 
> Subject: Re: PostgreSQL vs MySQL Throughdown! Teams? 
> To: [hidden email] 
> Message-ID: <[hidden email]> 
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed 
> 
> [hidden email] wrote: 
>> It is beginning to sound like people have a few things to say about the 
>> matter. There are lots of benchmarks out there and no one knows what to 
>> trust, so here's my proposal. 
>> 
>> We have a MySQL team and a PostgreSQL team. The two teams will develop a 
>> single benchmark that will function exactly the same on both systems. 
>> Something with a lot of concurrency and a balance between complex and 
>> simple queries. 
> 
> It's already been done, sort of. 
> 
> On the SPECjAppServer2004 benchmark, a test of J2EE performance, a 
> server running Postgres 8.2 scored 778 JOPS (jAppServer Operations Per 
> Second), and a server running MySQL 5 scored 721 JOPS.  But the hardware 
> isn't exactly comparable: the database server for Postgres was an 
> eight-core Sun Fire T2000 with 16 GB of memory, while the database 
> server for MySQL was a four-core Sun Fire X4100 with 8 GB of memory. 
> 
> http://www.spec.org/osg/jAppServer2004/results/res2007q3/jAppServer2004-20070606-00065.html
> http://www.spec.org/osg/jAppServer2004/results/res2007q2/jAppServer2004-20070411-00063.html
> 
> The TPC (Transaction Processing Council) also has a number of benchmarks 
> for various aspects of database performance.  I can't find any official 
> *recent* results from them for Postgres or MySQL, but this page has an 
> interesting comparison of the results from various subtests of TCP-D, 
> which measures query optimization: 
> 
> http://www.openlinksw.com/dataspace/vdb/weblog/vdb's%20BLOG%20%5B136%5D/1132
> 
> There are lots of different things people use RDBMS's for, so you can't 
> compose one benchmark and say it is *the* measurement for comparison. 
> 
> Here's an interesting essay by a PostgreSQL developer regarding the 
> strengths and weaknesses of each platform: 
> 
> http://www.postgresql.org/docs/techdocs.83
>