Does it matter where the binaries are?
- From: stephen booth <stephenbooth.uk@xxxxxxxxx>
- To: Oracle-L <oracle-l@xxxxxxxxxxxxx>
- Date: Thu, 10 Mar 2005 16:40:33 +0000
We have a collection of systems proposed where a set (I'm
delibaerately avoiding the word 'cluster' here as we're not talking
RAC) of Sun V440 boxes will be running Oracle instances attached to a
NetApp 920c filer cluster providing storage. Each server will be
running one live production database and one hot standby (dataguard)
of one of the database for which the live database is on one of the
other servers.
The datafiles for each database will be on the filer, each database
will be uniquely named to ensure no name clashes.
It has been suggested that we put the Oracle binaries on the filer so
when we patch one server we're patching them all. From discussions so
far it sounds like $ORACLE_BASE will point to a location on the server
but $ORACLE_HOME will point to a location on the filer (alert logs,
tnanames &c will be on the server storage, datafiles and binaries will
be on the filer).
Obviously we'll have to make sure trhat the Oracle user UID numbers
are the same accross all servers (but we have to do that anyway
regfardless of where the binaries live) and will have to ensure that
the correct permissions are set and the files in /usr/local/bin and
/var/opt/oracle are created (presumably riunnign the root.sh and
orainstRoot.sh scripts created when we do the first install will
accomplish this).
Could someone please tell me. Has anyone done this (seems a pretty
obvious thing to try)? Did it work? Is there anything we need to
watch out for (have I missed anything)? Is anyone aware of any
documentation on how to do this?
Thanks
Stephen
--
It's better to ask a silly question than to make a silly assumption.
--
//www.freelists.org/webpage/oracle-l
Other related posts:
