Quick note to myself that it is worth following the actual directions for setting up cron over at http://drupal.org/cron, instead of just a rough approximation. I have been doing
30 * * * * /usr/local/bin/wget http://my.site.com/cron.php >& /dev/null And have always been a little annoyed at the accumulation of cron.php.X files in my home directory, but never annoyed enough to figure it out.
Turns out wget really wants to save a file with the web content. Sending the content to /dev/null doesn't stop it. wget just happily creates an empty file. That is why the drupal.org docs helpfully suggest:
30 * * * * /usr/local/bin/wget -O - -q -t 1 http://my.site.com/cron.php-O - = Don't create a new file, instead send results to to standard output.
-q = quiet. This turns off the standard output (clever, eh?).
-t 1 = Only try once.
I've been growing more fond of CURL recently
The default action is to print to stdout, and it's also available in PHP as a library.
Thanks for pointing this
Thanks for pointing this out. I too have fallen victim to not being annoyed enough to figure out the correct way to hit the cron.php file. I always wondered why sending the download to /dev/null wasn't working, but never took the time to look. Thanks for taking the time to blog about it.