Dtrace is one of the coolest things in Solaris 10, adding great observability to applications. A Dtrace provider for PHP that adds probes to function entry and exit points has long been available. I finally got around to integrating this with the php in Cool Stack. 

Bryan Cantrill has some excellent examples of how to use dtrace to trace through your php code, down to the system libraries and kernel (if you want to go that far !) .

Instructions for installing the php dtrace extension

1. Download the shared library for the extension :

    dtrace.so.x86.bz2

…

We now have Developer support for Cool Stack (or Solaris + AMP).  This dedicated online support service for developers provides technical
assistance for code support, diagnostic advice, and programming
questions that may include:

• Sanity Checks
• Code level support
• Best practice guidance
• Workarounds when available
• And other forms of technical assistance

Sun Developer Expert Assistance Service is available to all developers,
with a cost of $49 (USD) per request or unlimited requests for an
annual subscription cost of $249 (USD).

Check out the details at http://developers.sun.com/services/expertassistance/

  …

I attended the Web 2.0 Expo held in San Francisco between April 15-18. Here are some thoughts on some of the things that caught my attention.

 Rich Internet Applications

There was a lot of talk about RIAs, many products ranging from full-blown development environments to languages, language environments etc. Some examples include Apollo (new announcement from Adobe), Django, curl (no - this is not the OSS libcurl), Silverlight (Microsoft's response to Adobe !)
So it seems that the applications are once again moving onto the desktop - after moving from fat client to the web, we have now realized that the web environment really is not rich enough for many applications. …

Sun just post SpecjAppserver2004 with a very excellent result of 712.87 SPECjAppserver 2004 JOPS@Standard using the latest SJSAS 9.0 Platform Edition on Sun Fire X4100 Cluster with MySQL 5.0.20 yesterday(Also check out Tom Daly's blog, Scott Oak's blog and Robert Lee's blog for more information)

SPECjAppserver is industry standard benchmark for performance testing J2EE application server. As part of important component of SPECjAppserver testing, database's performance and scalability also greatly affect the overall performance result in …

Sun just post SpecjAppserver2004 with a very excellent result of 712.87 SPECjAppserver 2004 JOPS@Standard using the latest SJSAS 9.0 Platform Edition on Sun Fire X4100 Cluster with MySQL 5.0.20 yesterday(Also check out Tom Daly's blog, Scott Oak's blog and Robert Lee's blog for more information)

SPECjAppserver is industry standard benchmark for performance testing J2EE application server. As part of important component of SPECjAppserver testing, database's performance and scalability also greatly affect the overall performance result in …

Sun just post SpecjAppserver2004 with a very excellent result of 712.87 SPECjAppserver 2004 JOPS@Standard using the latest SJSAS 9.0 Platform Edition on Sun Fire X4100 Cluster with MySQL 5.0.20 yesterday(Also check out Tom Daly's blog, Scott Oak's blog and Robert Lee's blog for more information)

SPECjAppserver is industry standard benchmark for performance testing J2EE application server. As part of important component of SPECjAppserver testing, database's performance and scalability also greatly affect the overall performance result in …

Hey, look! Peter mentioned me in his presentation at the UC this year.

UC2006-MySQL-Performance-Landscape.pdf

I let him use the Sunfire T2000 I borrowed from Sun (which I should return today) to generate some statistics for the talk.

© cjcollier for C.J.'s WordPress of studlyness, 2006. | Permalink | No comment

Add to del.icio.us

Search blogs linking this post with …

If you have been reading my previous entries the answer you will think is ‘not bloody well’.

After about 3 days of tuning we doubled the throughput, and got a much nicer picture, outperforming a x86-64 machine by 2.5 times in one case.

Thanks to Luojia Chen (Jenny) from Sun, Peter Zaitsev from Mysql, and Colm MacCárthaigh & Mads Toftum from the ASF.

oh.. the benchmark.. I nearly forgot ;-)

(Oh people..please link to the blog entry, and not the paper itself.. Thanks)

update: people were having issues downloading the PDF. so I placed a mirror of it here

this is the output of dtrace -n 'pid$target:::entry{ @[probefunc] = count() }'

when I run mysqlslap -c 20 -i 20

  pthread_getspecific                                             365
  malloc_internal                                                 377
  my_malloc                                                       377
  free                                                            378
  malloc                                                          378
  my_wc_mb_filename                                               399
  my_no_flags_free                                                420
  strmake                                                         444
  pthread_self                                                    468
  memset                                                          582
  alloc_root                                                      630
  strmov …

ok.. now this is getting stranger.

I have 2 windows open, in one I get

$ mysqlslap --use-threads -psunfire -c 90 -i 500
Benchmark
        Average number of seconds to run all queries: 2.180 seconds
        Minimum number of seconds to run all queries: 0.779 seconds
        Maximum number of seconds to run all queries: 4.497 seconds
        Number of clients running queries: 90
        Average number of queries per client: 0

and in the other i get

$ mysqlslap --use-threads -psunfire -c 90 -i 500
Benchmark
        Average number of seconds to run all queries: 0.012 seconds
        Minimum number of seconds to run all queries: 0.002 seconds
        Maximum number of seconds to run all queries: 0.015 seconds
        Number of clients running queries: 90
        Average number of queries per client: 0

same machine, same user-id, same binary. the only difference is that mysql started in the ‘fast’ window.