For some reason when my CRON runs for my Oracle id, the environment variables don't appear to get passed. As an example, if I log in as Oracle and run a script that does the following

echo $PATH > path.txt

then I am returned the following:

/usr/local/bin:/oracle/product/9.2.0.6/bin:home/oracle/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/opt/gnome/bin:/opt/kde3/bin

But...

When I set up cron for Oracle, then the script does the following (output from echo $PATH

/usr/bin:/bin

Why isn't cron picking up Oracle's correct PATH? The script isn't passing my correct env either. Any ideas?

Thanks in advance,

Bret
SLC, UT

Depending on the version of cron that you run, it probably has a cut down environment to run your jobs in. Dillon's cron for example, only keeps the USER, HOME, and SHELL variables when running a user's job.

Using the full path to the commands used in the script and exporting the required environment variables in the script should allow it to run.

Hi Steve,

Thanks for the update. I'll try it.

We have many scripts that will need to be ported from HP-UX. The problem is that we don't want to modify every script with the environment variable information that they each need. Obviously, the cron env is different from shell environment of the oracle user (or any user for that matter). Why is this, and how do I make the cron env to be the same as the user's shell env?

As an example...
If I do an 'env' in cron, I only get:
SHELL
PATH
PWD
HOME
SHLVL
LOGNAME
_

When I do an `env` in my shell, I get about 70 variables. Is there any way to get these variables visible in my cron? I can't get any of my scripts to run in cron. Thanks.

Bret Goodfellow
SLC, UT

Unfortunately, not with the version of cron I'm using. It may be possible to create a file that holds your environment variables and source that at the top of each script. It does mean modifying each script still, but you could do it with sed.