![]() | |
![]() |
| | Thread Tools | Display Modes |
#11
| |||
| |||
|
|
Our failover cluster setup is as follows: -2 nodes, each: 4 QuadCore CPUs (total of 16 shown in Task Mgr), 64 GB RAM -Windows Server 2003 Enterprise x64 Edition, SP2 -single named instance: SQL Server 2005 64-bit, SP2 This cluster has been set up in a test environment (prior to deploying to production in a month or so). While running some load tests, despite the CPU cranking up to 100%, we noticed the max memory usage (Total Server Memory:8,438,016) was way less than what we expected (Target Server Memory:60,245,744). Being that this is a single instance node we didn't jack with the sp_configure max/min server memory settings. We even did cross joins with some of our bigger tables (30 million rows+) to ensure memory intensive queries were being run. The SQL Server Error Log shows that it is indeed using AWE to allocate memory (after having taken care of the obligatory "Lock pages in memory setting"). My research to date has not helped clarify if there's any additional tweakings that I need to make to get SQL Server to use more memory. Is anybody aware of something else that I must do to get the Total Server Memory to be more inline with the Target Server Memory? Any help will be most appreciated. Raul |
![]() |
| Thread Tools | |
| Display Modes | |
| |