After Zammad upgrade nginx starts instead of apache2

Technical assistance
1

Infos:

  • Used Zammad version: 6.4.1-1733418007.181ec27f.jammy
  • Used Zammad installation type: package
  • Operating system: Linux Mint 21.3 Virginia
  • Browser + version: Chrome 131.0.6778.70

Hi,

we use Zammad together with Apache2 because of SSO via Kerberos. Every time after system and Zammad upgrade via “apt update && apt upgrade” the system starts nginx instead of apache2.

Each time I have to do the following steps manually

  • sudo systemctl stop nginx
  • sudo systemctl disable nginx
  • sudo systemctl start apache2
  • sudo systemctl enable apache2

Of course I’s not a big effort, but I would like to avoid it. Where can I change this behaviour?

Thanks :slight_smile:

2

You’re using a Ubuntu package on Linux Mint which is not part of our officially supported distributions.

3

Hi.

Can confirm this also happens on current Debian 11. Just updated to 6.4.1-1736159868.df374f06.bullseye.

Like OP said, the problem first occurred after we have configured SSO for domain clients using Kerberos.

This happened on every update using the official package source of Zammad (which for me is: https://dl.packager.io/srv/deb/zammad/zammad/stable/debian 11 main)

4

Well remove nginx in favor for apache2 then. Having two web servers installed while one is set to inactive isn’t really needed at all.

5

Well… yeah. :no_mouth:
I could have found that out myself if I had checked the dependencies of the Zammad package and stumbled across the ‘apache2|nginx’…

Thanks for the hint and nevermind…

6

Thanks, after removing nginx the problem is solved (sudo apt purge nginx).
I recently did two updates/upgrades via apt without problems.

7

Should have looked into the postinstall code before answering.
Though weeks on my end.

Any way, the reason an update strikes your back like that without removing nginx (if you prefer apache2) is that our postinstall scripts prefers nginx during the webserver detection and thus, from your POV, starts the wrong web server back online. So much for the reasons.

8

Thanks for explaining!

Well, I think it is good as is, because it will fit most people’s setup. The only alternative method, that comes to my mind, would be to just leave the web servers autostart settings as is, in case apache2 and nginx are both installed on the system.

However, in my case I had just forgotten to remove nginx after implementing SSO. I think this „issue“ will mainly occur in setups like mine.

Maybe it would be useful to just place a small hint in the setup docs for SSO. (Not sure if this is the only part in your docs, where you recommend to switch to apache2).

9

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.