Comments on: Benchmark MySQL Proxy and HSCALE http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/ Thu, 28 Aug 2008 23:05:55 +0000 http://wordpress.org/?v=2.3.3 By: Robert Hodges: Myosotis Connector: a Fast SQL Proxy for MySQL and PostgreSQL | Maildir.net Mailling list archive http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-4107 Robert Hodges: Myosotis Connector: a Fast SQL Proxy for MySQL and PostgreSQL | Maildir.net Mailling list archive Mon, 14 Jul 2008 17:48:09 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-4107 [...] As other people have discovered, proxying is very CPU-intensive. It also involves a lot of concurrency, since a proxy may have to manage hundreds or even thousands of connections. Java is already fast in single threads–after a few runs through method invocations, the JVM has compiled the bytecodes down to native machine code. In addition, Java uses multiple CPUs relatively efficiently. Myosotis uses a thread per connection. Java automatically schedules these on all CPUs and optimizes of memory access in multi-core environment. [...] […] As other people have discovered, proxying is very CPU-intensive. It also involves a lot of concurrency, since a proxy may have to manage hundreds or even thousands of connections. Java is already fast in single threads–after a few runs through method invocations, the JVM has compiled the bytecodes down to native machine code. In addition, Java uses multiple CPUs relatively efficiently. Myosotis uses a thread per connection. Java automatically schedules these on all CPUs and optimizes of memory access in multi-core environment. […]

]]> By: pero http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1274 pero Mon, 05 May 2008 19:28:39 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1274 Jan, I re-ran the tests with the current svn version. There are some great improvements to be seen! Take a look at the results here: http://pero.blogs.aprilmayjune.org/2008/05/05/update-benchmark-hscale-with-mysql-proxy-070-svn-against-061/ Cheers, Peter Jan,

I re-ran the tests with the current svn version. There are some great improvements to be seen! Take a look at the results here:

http://pero.blogs.aprilmayjune.org/2008/05/05/update-benchmark-hscale-with-mysql-proxy-070-svn-against-061/

Cheers,

Peter

]]> By: Jan Kneschke http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1259 Jan Kneschke Mon, 05 May 2008 18:09:29 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1259 You may want to give the SVN version of MySQL Proxy a try. It reduces the caches the compiled lua script and should reduce the overhead of hscale. You may want to give the SVN version of MySQL Proxy a try. It reduces the caches the compiled lua script and should reduce the overhead of hscale.

]]> By: pero http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1237 pero Mon, 05 May 2008 16:34:00 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1237 Brian, mysqlslap was really a great help here. And it worked like a charm! As minor improvements: it would be great to get the result for *each* run (to correlate it to vmstat or sar data) and to be able to use multiple statements. Greetings, Peter Brian,

mysqlslap was really a great help here. And it worked like a charm!
As minor improvements: it would be great to get the result for *each* run (to correlate it to vmstat or sar data) and to be able to use multiple statements.

Greetings,

Peter

]]> By: Brian Aker http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1232 Brian Aker Mon, 05 May 2008 16:20:45 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1232 Hi! Any feedback on Mysqlslap? (positive or negative) Cheers, -Brian Hi!

Any feedback on Mysqlslap? (positive or negative)

Cheers,
-Brian

]]> By: Kay Roepke http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1224 Kay Roepke Mon, 05 May 2008 15:54:51 +0000 http://pero.blogs.aprilmayjune.org/2008/05/05/benchmark-mysql-proxy-and-hscale/#comment-1224 Hi! Regarding the final thought #2: Yes, the proxy currently is single-threaded. We are aware of this limitation and it will change in some way in a future release to make proper use of current multi-core/SMP systems. cheers, -k Hi!

Regarding the final thought #2:
Yes, the proxy currently is single-threaded. We are aware of this limitation and it will change in some way in a future release to make proper use of current multi-core/SMP systems.

cheers,
-k

]]>