Websites down after upgrade
Last night I ran an upgrade of my server (Debian trixie/sid), and afterwards none of my websites are working. Server version: Apache/2.4.57 (Debian) The websites are a mix of static pages and Wordpress blogs - all Virtual Hosts. None of them resolve, and I am getting "unable to connect" messages - but nothing is showing up in the /var/log/apache2/error.log or the access.log. Apache is running - I can see the restarts in the error log - but nothing is happening when I try to hit a page. Does this ring any bells for anyone? Is there a new configuration wrinkle that I didn't adapt to? Thanks for any help you can provide. Wililam
William Witteman via talk wrote on 2023-09-15 05:47:
The websites are a mix of static pages and Wordpress blogs - all Virtual Hosts.
None of them resolve, and I am getting "unable to connect" messages - but nothing is showing up in the /var/log/apache2/error.log or the access.log.
Do the vhosts have separate log files? If so, anything in those indicating an issue?
Apache is running - I can see the restarts in the error log - but nothing is happening when I try to hit a page.
Anything revealed by running `curl --head $site` `curl --head --location $site` ?
Does this ring any bells for anyone? Is there a new configuration wrinkle that I didn't adapt to?
I don't run Debian, it doesn't ring any bells, but I thought I'd throw a couple ideas out there to see if it helps at all. rb
Thank for your the hint! I think that the HTTPS redirect isn't working. The curl --head $site gives me this: HTTP/1.1 301 Moved Permanently But it doesn't connect to port 443. sudo netstat -ntupl | grep :443 doesn't show anything. Perhaps I am missing a Listen 443 directive? Similarly, if I hit the bare IP address I get the default virtual host, and if I "lynx localhost" I see the same, so Apache and PHP are definitely in play. On Fri, 15 Sept 2023 at 09:04, BCLUG via talk <talk@gtalug.org> wrote:
William Witteman via talk wrote on 2023-09-15 05:47:
The websites are a mix of static pages and Wordpress blogs - all Virtual Hosts.
None of them resolve, and I am getting "unable to connect" messages - but nothing is showing up in the /var/log/apache2/error.log or the access.log.
Do the vhosts have separate log files? If so, anything in those indicating an issue?
Apache is running - I can see the restarts in the error log - but nothing is happening when I try to hit a page.
Anything revealed by running
`curl --head $site` `curl --head --location $site`
?
Does this ring any bells for anyone? Is there a new configuration wrinkle that I didn't adapt to?
I don't run Debian, it doesn't ring any bells, but I thought I'd throw a couple ideas out there to see if it helps at all.
rb
--- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
FIXED! I looked at /etc/apache2/ports.conf, and the "Listen 443" directives are wrapped in a conditional that requires the ssl module. So I ran sudo a2enmod ssl ...and then sudo systemctl restart apache2 Now all my sites are back. Thank you! On Fri, 15 Sept 2023 at 09:56, William Witteman <wwitteman@gmail.com> wrote:
Thank for your the hint!
I think that the HTTPS redirect isn't working.
The curl --head $site gives me this: HTTP/1.1 301 Moved Permanently
But it doesn't connect to port 443.
sudo netstat -ntupl | grep :443 doesn't show anything. Perhaps I am missing a Listen 443 directive?
Similarly, if I hit the bare IP address I get the default virtual host, and if I "lynx localhost" I see the same, so Apache and PHP are definitely in play.
On Fri, 15 Sept 2023 at 09:04, BCLUG via talk <talk@gtalug.org> wrote:
William Witteman via talk wrote on 2023-09-15 05:47:
The websites are a mix of static pages and Wordpress blogs - all Virtual Hosts.
None of them resolve, and I am getting "unable to connect" messages - but nothing is showing up in the /var/log/apache2/error.log or the access.log.
Do the vhosts have separate log files? If so, anything in those indicating an issue?
Apache is running - I can see the restarts in the error log - but nothing is happening when I try to hit a page.
Anything revealed by running
`curl --head $site` `curl --head --location $site`
?
Does this ring any bells for anyone? Is there a new configuration wrinkle that I didn't adapt to?
I don't run Debian, it doesn't ring any bells, but I thought I'd throw a couple ideas out there to see if it helps at all.
rb
--- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
William Witteman wrote on 2023-09-15 07:03:
FIXED!
I looked at /etc/apache2/ports.conf, and the "Listen 443" directives are wrapped in a conditional that requires the ssl module.
So I ran sudo a2enmod ssl ...and then sudo systemctl restart apache2
Now all my sites are back. Thank you!
That's awesome, congrats on having everything back up! rb
That begs a question... how was it before the upgrade? -- On 2023-09-15 10:03, William Witteman via talk wrote:
FIXED!
I looked at /etc/apache2/ports.conf, and the "Listen 443" directives are wrapped in a conditional that requires the ssl module.
So I ran sudo a2enmod ssl ...and then sudo systemctl restart apache2
Now all my sites are back. Thank you!
Everything was fine. Something went wrong during the upgrade and one of the troubleshooting things I tried was to uninstall and reinstall apache. I didn't realize that that disabled all of the sites and modules. On Fri, Sept 15, 2023, 10:40 William Park via talk <talk@gtalug.org> wrote:
That begs a question... how was it before the upgrade? --
On 2023-09-15 10:03, William Witteman via talk wrote:
FIXED!
I looked at /etc/apache2/ports.conf, and the "Listen 443" directives are wrapped in a conditional that requires the ssl module.
So I ran sudo a2enmod ssl ...and then sudo systemctl restart apache2
Now all my sites are back. Thank you!
Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
participants (3)
-
BCLUG -
William Park -
William Witteman