[hurin]

Hi,

I just did a script in the purpose to save all directory from a specific one's (/var/www/html ).
The script works perfectly when launched manually, but not at all when it's the cron that launch it automatically =/

here's the script =>

Code:
for SITE in `ls $MYDIR | grep -v "chroot\|default"` ; do
if [ -d $MYDIR${SITE} ]; then
echo 'here, before tar '$MYDIR${SITE} >> log.txt
tar -czvf ./${SITE}.tar.gz $MYDIR${SITE}
fi
done


I just have no idea what's going on, i checked into the log.txt, the path and the directory name are good, but, it looks like only one directory has been zipped, and nothing else happen,...

Here's what i got from the log =>
Quote:
here, before tar /var/www/html/my_sitename


Hoping someone could give me a hint on this issue,

Hurin

PS: sorry for my low english level ^^

[SimonJM]

A couple of things spring to mind - the environment in use (there are differences between your user environment and the one cron providesn - maily in PATH but also in what shell is used). Where does the variable $MYDIR come from?

[hurin]

Ok, i just found out how to make it work, even if i don't understand why, but the fact to redirect the output of the command tar -czvf has been enough to make it work 0_o
So, before, i had this =>
tar -czvf ./${SITE}.tar.gz $MYDIR${SITE}

And now, i got this =>
tar -czvf ./${SITE}.tar.gz $MYDIR${SITE} >> log.txt 2>&1

So, if anyone know why it's working now, i would gladly learn it from him

[SimonJM]

No idea why appending the stdout and stderr from the tar command is now making the (I presume) loop process more directories! What happens if you run the revised script manually?

<edit>
Now, having tried the code (and giving MYDIR a suitable value), I can really see no reason as the original code works fine for me - both run manually and via cron.
</edit>