Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: strange sqlplus problem w/ Oracle 7.3.4

Re: strange sqlplus problem w/ Oracle 7.3.4

From: Sasha Ostojic <sasha_at_UltimateTech.com>
Date: Mon, 11 Jan 1999 15:22:49 -0800
Message-ID: <77e16p$s9q$1@news-sj-3.cisco.com>


Actually, sqlplus evaluates env variables just fine. In fact, the SPOOL command works just fine as long as the specified path expands to less than 65 characters. This all worked fine in Oracle 7.3.2, and when one of our customers upgraded to 7.3.4 all of a sudden they got an error in one of our scripts. I couldn't figure it out until I noticed that it was failing only on long path names.

I installed the 7.3.4.3 patch, but that still didn't fix it. I haven't had a chance to call Oracle yet, but I do need a fix because all these scripts are on CDs that have already been manufactured.

.sasha

Stephen Harris wrote in message <7787d0$e9n$4_at_nebula.mpn.com>...
>
>SQL*Plus doesn't evaluate environment variables like this, so you
>are trying to open a file literally called '$DBPATH/....'.
>
>What confuses people is things like
> sqlplus user/pass << EOF
> SPOOL $DBPATH/....
> EOF
>works. However this is because the shell is evaluating the variable
>before SQL*Plus sees it, and so SQL*Plus actually sees the full pathname.
>
>The SIGSEGV is a bug (IMHO) - bad error trapping in the SQL*Plus program;
>I assume the code isn't testing the result of the fopen() [ or open() - I
>haven't bother to trace this ] and is blindly writing to an unopened
>file descriptor. I've seen this once before but didn't think it worth
>raising a TAR over :-)
>
>--
>
>rgds
>Stephen
Received on Mon Jan 11 1999 - 17:22:49 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US