Here are a couple suggestions:

1. Use absolute paths in cron: cron jobs are executed by crond in a minimal environment without a shell (unless you force one to be created), the regular shell setup (including things like your PATH variable that helps map curl to /usr/bin/curl) never happens. More reference

2. Do you intend to log the entire response HTML? I think right now, you are saving everything that comes back from http://url.com/cron/email in /var/log/cron. If your response is just a short plaintext thing that should be fine, but if it is returning html there’s going to be a lot of junk in your logs. Add the --silent option to avoid that.

3. Pipe output to dev/null The Cron daemon that runs your command wants to send the output to the email address associated with your account, but I am guessing you don’t have email configured and you dont need that so add >/dev/null 2>&1 to the end of the command to pipe all output to /dev/null (same as ignoring it)

So with all that, try:

* 17 * * * /usr/bin/curl --silent http://url.com/cron/email >/dev/null 2>&1

Hey @aha

Thank you for answering my question.

Unfortunately it’s not working. I’ve tried to add the line you’ve recommended, and made it run every minute * * * * * just for testing purpose, but nothing seems to work.

I have no intention of logging anything. I just want to run the URL once a day at 5pm - nothing else. Since on that webpage it’s doing some logic. Maybe this is not even possible? To run a webpage like if you visited it in the browser?

One more tip for debugging: to see the logged results of the cron task you set up to run every minute, use grep CRON /var/log/syslog to print out all the lines related to cron. You should see something like:

CRON[2179]: (root) CMD (/usr/bin/curl http://url.com/cron/email)

However - you are right, there is a big difference between running curl http://url.com/cron/email and visiting http://url.com/cron/email in the browser:

When you visit http://url.com/cron/email in the browser, if it returns an HTML doctype, the browser will: A) Download the HTML, B) Download any Javascript, CSS, Images referenced in the HTML, C)Render the page, D)Execute all Javascript.

When you curl http://url.com/cron/email, it is just A) Downloading the HTML

So if the thing that causes the email send is executed via javascript on the page, curling it will not help.