Since my recent move to the Mac running OS X I have learnt some new stuff about PHP and Apache the hard way. OS X is shipped with Apache and PHP by default and if you compile Apache and PHP like you normally do under Linux you quickly get surprised. When I compile Apache I always use a prefix so I have it installed under /usr/local/apache or something like that. I like to have it installed like this so I can quickly change between Apache and PHP versions and configurations. Under OS X you quickly notice that Apache is not completely installed in the directory you added as prefix. The reason for that is that Apache has a different installation layout under OS X. So, you simply have to use the configure parameter --with-layout to make things work like you are used to.

The compilation switches I use are:

cd apache_1.3.34
make clean
./configure --prefix=/usr/local/apache \
                    --enable-module=so \
                    --enable-module=rewrite \
                    --with-layout=Apache
make
sudo make install
cd ..
cd php-4.4.2
make clean
./configure --prefix=/usr/local/apache  \
            --enable-mbstring \
            --with-apxs=/usr/local/apache/bin/apxs \
            --enable-cli \
            --enable-exif \
            --with-mysql \
            --with-mbstring
make
sudo make install 

I did some benchmarks on how eZ publish performs on the new Intel based MacBook Pro running OS X version 10.4.6. The PowerPC based Mac machines has really poor performance with eZ publish when I have tested or heard about tests on it, so I was excited about the results.

I did a plain installation of eZ publish 3.7.5 and did a simple ab -c 5 -n 50 and got 12.78 pages pr second. It also feels quick on normal operations in eZ publish. The new Intel based Macs are now a good option for doing eZ publish development. It is still slower than Linux, but for development and testing it is good enough.

I also did some tests with the apc.stat=0, but I did not see any performance improvement. I was hoping that this new setting would remove the stat overhead found on FreeBSD based systems, but atleast when I tested I saw no difference.

The configuration of the laptop is 1.83 GHz and 1GB ram. Running Apache 1.3.34, PHP 4.4.2 and APC 3.0.10.

Since the testing of MySQL compiled with Intel ICC showed good numbers I carried on some testing with Apache, PHP and APC also compiled with the ICC compiler. The test was done with eZ publish running on a Linux server with Xeon processors.

All tests showed speed improvements. When I ran tests on template processing, which is the most PHP intensive operations in eZ publish, it showed improvements from 17.4% to 19.6% faster with the ICC compiler. Total script runtime measured with ab ranged from 6.3% to 12.7% faster. Of course these numbers are a bit lower since they include I/O like network transfer of the pages and so on. The more PHP intensive and heavy the script was the faster it ran under ICC.

Of course this compiler comes with a $399 pricetag so you have to compare that with the performance gained with it. In our case it looks like heavy operations are boosted by close to 20%.

To get your stack compiled with ICC you simply install the compiler as instructed by Intel and export the CC environment variable.

export CC=/opt/intel/cc/9.0/bin/icc 

Once this is set up you can do a make clean and rm config.cache followed by a ./configure. For APC you also need to run phpize before you configure. Then you make and make install like normal. No tweaking was done other than this so there is mabye room for further tweaking of compiler parameters.

It would be good to hear other people's experience with PHP and the ICC compiler.

A couple of pictures during my trip to Vancouver earlier this year is now uploaded in the photo gallery.

I've done some benchmarking today with eZ publish CMS and MySQL running a process inserting articles (objects) into the CMS system. I ran the test on both the standard GCC compiled version of MySQL and the Intel ICC compiled version. The total process was 11.9% faster and the MySQL part was 20.2% faster. So atleast on heavy write operations the ICC compiled version of MySQL is running faster with eZ publish. This is good news for people running write intensive eZ publish sites. When testing with read queries I did not see that much of a difference, I will do more tests on heavy read intensive operations to find where we can gain performance also here.

I will be speaking at the MySQL Users Conference 2006 in Santa Clara on Wednesday the 26th of April.The topic of the session is Web Application Clustering with MySQL. Zak Greant will be co-presenting with me.

I just enabled the photo gallery. You can browse some images I made on a trip to Iquitos in Peru last October. More photos are coming soon.

We've just received, installed and configured the new cluster of servers to use for eZ publish performance testing. It's the hardware we will use to test and tweak the new clustering functionality in eZ publish 3.8. It's 6 servers in total, all dual Xeon 3.2 GHz with from 1-8GB of ram in them. The picture shows Sten with the rack.

Oh, the servers will be demo'ed on the eZ publish conference for those of you attending.