I tested both platforms on a SAN every day for 26 days 18 hours a day straight.
I came to this conclusion. The kernel does make a huge difference in the two platforms. Especially on the chipset patches and the disk IO subsystem. If you go with AMD do not use 2.6.7 or ext3. EXT3 has a bug in it and its use with O_DIRECT, and 2.6.6+ IO scheduler is not enterprise ready, although by description it sounds like it is. The VM system is to fast for the scheduler. Use 2.4 SUSE for AMD. If you need to run RedHat DO NOT RUN THE RedHat AS 3.0 ISO, run RedHat with SUSEs kernel in 64 bit mode. RedHat AS back ported some 2.6 features and the disk IO scheduler is NOT an "Advance" This is the fastest config that our team has come up with. It's nearly 3 times faster then a XEON with 4GB of ram and 5 times faster with 8GB of ram. As many of you know, Friendster was very slow, due to our AMD solution and some mySQL optimizations Friendster is FAST. DVP ---- Dathan Vance Pattishall http://www.friendster.com > -----Original Message----- > From: Miles Keaton [mailto:[EMAIL PROTECTED] > Sent: Tuesday, July 27, 2004 1:48 PM > To: [EMAIL PROTECTED] > Subject: tests comparing AMD64 to Xeon MySQL performance? > > We've got a really high-load MySQL server and are planning to get a new > server. > > Has anyone seen tests comparing performance of MySQL on AMD64 versus Xeon > CPUs? > > -- > MySQL General Mailing List > For list archives: http://lists.mysql.com/mysql > To unsubscribe: > http://lists.mysql.com/[EMAIL PROTECTED] -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
