Re: sqlplus on HP-UX
Date: Mon, 24 Aug 2009 10:04:21 -0800
That's quite interesting...on the Tru64 machine, it looks like this is completely disabled,
> oracle_at_wrangell> /usr/sbin/chatr ~oracle/product/rel102040/bin/sqlplus
> current values:
> 64-bit COFF executable
> execute from data: disabled
> use big pages: default
but on the HP-UX machine, my output looks similar to your example.
Joel, regarding the docs, I'll look more carefully, but at a first glance through the installation manual, the only difference looks like for entries in the listener.ora file. I'll check the Solaris documents to see what the difference there is to make sure I'm looking in the right place.
Thanks for the suggestions/comments.
Adams, Matthew (GE Indust, ConsInd) wrote:
> Try running 'chatr sqlplus' in both environments and examing the top
> few lines to see what happens if the LD_LIBRARY_PATH is not set. If
> SHLIB_PATH searching is enabled, see what the value of that
> environmental variable is.
> $ chatr sqlplus
> 64-bit ELF executable
> shared library dynamic path search:
> LD_LIBRARY_PATH enabled first
> SHLIB_PATH enabled second
> embedded path enabled third
> -----Original Message-----
> From: oracle-l-bounce_at_freelists.org
> [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Maureen English
> Sent: Monday, August 24, 2009 12:55 PM
> To: oracle-l_at_freelists.org
> Subject: sqlplus on HP-UX
> We have Oracle release 10.2.0.3.0 and 10.2.0.4.0 on our HP-UX
> and HP Tru64 machines. I found something very strange, though,
> on the HP-UX machines.
> On the HP-UX machines, if our LD_LIBRARY_PATH is set to the
> 10.2.0.3.0/lib, when we start sqlplus, even though ORACLE_HOME
> and PATH variables are correctly pointing to 10.2.0.4.0, the
> version of sqlplus that is started is the 10.2.0.3.0 version.
> If LD_LIBRARY_PATH is not set at all, or is set to 10.2.0.4.0/lib,
> then the correct version of sqlplus is started.
> On the HP Tru64 machines, the correct version of sqlplus is
> started no matter what we set LD_LIBRARY_PATH to.
> I do know that we really should set LD_LIBRARY_PATH in all cases,
> but due to an oversight, it wasn't being set in our database
> startup scripts. So, now I'm just curious as to why we are now
> seeing a different behavior.
> Any comments?
> - Maureen