BTW, this uses SCAN listener and endpoints all owned by grid and running off the GI home. Alan.- On Mon, Oct 15, 2012 at 5:23 PM, Guillermo Alan Bort <cicciuxdba@xxxxxxxxx>wrote: > Hi, > > I have the following scenario: > OS: AIX 6.1 > RDBMS: Oracle 11.2.0.3 CPU JUL2012 (no other patches, the PSU failed due > to lack of space) > This is a 2 node RAC. > Separate users for GI (grid) and RDBMS (oracle). > > So far this is pretty standard. > > The pmon starts correctly: > dbserver01:/u1/app/grid/11.2.0/network/admin>ps -ef |grep pmon > oracle 26345656 1 0 Oct 13 - 0:57 ora_pmon_<SID>1 > grid 28704882 1 0 Oct 12 - 1:08 asm_pmon_+ASM1 > > however, the server processes are started with grid: > dbserver01:/u1/app/grid/11.2.0/network/admin>ps -ef |grep oracle<SID> > *grid *6291568 1 0 Oct 13 - 0:00 oracle<SID>1 > (LOCAL=NO) > *oracle *22937652 10354818 0 15:59:09 - 0:00 oracle<SID>1 > (DESCRIPTION=(LOCAL=YES)(ADDRESS=(PROTOCOL=beq))) > > (note that SID is the same always as there is only one DB in this RAC) > > Now. I've never seen this before, maybe I wasn't paying attention or maybe > this is something exclusive to AIX. But I find this odd. Has anyone seen > this before? is this expected behaviour or a wrong config somewhere? Or > maybe a bug? > > Thanks > > Alan.- > > -- //www.freelists.org/webpage/oracle-l