I'm on a shared hosting solution that I just moved over to. After the move, I did an upgrade to the newest CiviCRM version but left WordPress intact (no upgrade). Scheduled cron jobs were working fine on the previous host server, but not anymore.
Current configuration: PHP: 7.2.13 CiviCRM: 5.8.2 WordPress 4.9.9
I updated the wget command I was using to capture the results in logs. In CPanel, the job looks like this: wget -O /home/xxx/public_html/cron_output.log -t 1 -a /home/xxx/public_html/cron.log 'https://xxx/wp-content/plugins/civicrm/civicrm/bin/cron.php?name=xxx&pass=xxx&key=xxx'
I know the syntax of the command is right because I can run the https: command from a web browser and everything works.
The cron.log shows that it's run with the 200 OK response. The cron_output.log is empty (should it be?).
--2019-01-02 12:50:01-- https://xxx/wp-content/plugins/civicrm/civicrm/bin/cron.php?name=xxx&pass=xxx&key=xxx Resolving xxx (xxx)... x.x.x.x Connecting to xxx (xxx)|x.x.x.x|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 0 [text/html] Saving to: ‘/home/xxx/public_html/cron_output.log’
The server logs that my hosting provider sent me show that the jobs run.
I've also scheduled the same job in easycron and it shows that's it's run successfully.
So I know that cron is running and the syntax is right. But the scheduled jobs in Civicrm don't run. There's nothing in any error_log anywhere that indicates there's a php problem. I've looked at the questions here, and checked that PHP and the MySQL database are on the same timezone, and they are.
I cannot figure out why Civicrm is not running the scheduled jobs. I've tested with a scheduled mailing that sits there and never gets delivered (until you run the https:// command line manually, then it completes). Also the Administer -> System Settings -> Scheduled Jobs show that they are not being completed unless the cron job is manually triggered. And Civicrm has the orange System Status: Warning: Cron Not Running
I'm stumped, and my provider says the issue is not with the server, and I tend to agree.
Any thoughts as to next steps?