Thanks- I saw this blog, but he does not have any performance conclusions on hugepages vs. 11g AMM for larger systems. On smaller, 32 bit systems there was no clear winner. -----Original Message----- From: D'Hooge Freek [mailto:Freek.DHooge@xxxxxxxxx] Sent: Wednesday, July 21, 2010 12:11 PM To: Crisler, Jon; oracle-l@xxxxxxxxxxxxx Subject: RE: 11g AMM tmpfs vs hugepages for best overall performance Jon, Yes, you can still pick between AMM and "normal" memory parameters. For the comparison, I suggest you look at the following blogpost of Kevin Closson: http://kevinclosson.wordpress.com/2007/08/23/oracle11g-automatic-memory-management-and-linux-hugepages-support/ Regards, Freek D'Hooge Uptime Oracle Database Administrator email: freek.dhooge@xxxxxxxxx tel +32(0)3 451 23 82 http://www.uptime.be disclaimer: www.uptime.be/disclaimer ________________________________________ From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Crisler, Jon Sent: woensdag 21 juli 2010 18:03 To: oracle-l@xxxxxxxxxxxxx Subject: RE: 11g AMM tmpfs vs hugepages for best overall performance And a follow up question- if I implement a large /tmpfs - shm, I can still use choose to not use AMM and use hugepages, correct ? From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Crisler, Jon Sent: Wednesday, July 21, 2010 11:57 AM To: oracle-l@xxxxxxxxxxxxx Subject: 11g AMM tmpfs vs hugepages for best overall performance Which is better for performance on large Red Hat 5 systems (64gb+ memory, 8+ cpu's) - using 10g style shm settings and hugepages, OR the newer 11g Automatic Memory Manager (which does not support hugepages). The system I am building is a 6 node 11g R1 RAC, memory somewhere between 64gb and 256gb (not sure yet), 8 cpu per node. This machine will support a huge workload. -- //www.freelists.org/webpage/oracle-l