After an update to Webmin today at 1100 central, 5 out of 6 virtual domains hosted on my box no longer parse PHP files. This has left us in a production down situation.
I have tried to pay Virtualmin support to help, but have only gotten questions about why I would be running PHP 8 and criticism of the base configuration of this server instead of working the problem (we remove mail related software because no mail services are needed for example).
Whenever any php file is rendered, including a file with exactly “hi” in the file with zero PHP code, the result is a 500 server error with this in the logs:
[cgi:error] [pid 17175] [client x.x.x.x:53959] End of script output before headers: php8.0.cgi
Comparing the 1 working site to the 5 broken sites, the 1 working site is set to use mod_php for the PHP execution method. The 5 broken sites are set to use the CGI wrapper.
I believe they should all be set to mod_php, and mod_php is installed and working for the 1 working site, but mod_php is not listed in the virtualmin UI as an option for those 5 broken sites so I cannot change it. mod_php IS listed and selected for the 1 working site.
My guess is that the webmin update tripped over the virtual host or apache configs and removed the mod_php options - but I do not know how that works, I’m only guessing based on the before and after that I’m seeing. This server has been configured how it is and running for almost a year with daily updates and weekly reboots.
Is there anyone listening who might be able to help? I can restore the server to last night’s backup, but when the update runs again, I believe this will simply break again and need to understand the root cause of the problem to prevent it from occurring again.
|OS type and version:||Ubuntu 18.04.5 LTS|
|Related products version:||Apache 2.4.48, PHP 8.0.10, MySQL 5.7.35|