[Technitium DNS] Running update to v15 breaks container

[frontend-report]

Script: Technitium DNS

Reported by: @nickmv5

Guidelines

• Read and understood guidelines: yes
• Verbose mode used: yes

Issue occurs during: Script Update

Environment

• Linux distribution: Debian
• Proxmox version: 9.1
• Default settings: No
• Advanced settings: No
• Exact command: bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/ct/technitiumdns.sh)"

Description

If you run "update" command on the CT to get to v15 of Technitium, the script fails to verify the version of the newest download, then proceeds to somehow corrupt the container into an unusable state, requiring a restore from backup.

Steps to reproduce

1. Run update

That's it. It'll error out within about 10 secs and leave you with an unusable instance of Technitium.

Error output

N/A

[CerberusStyle]

you are using an alternate dns server when trying to update ?

[CerberusStyle]

Yes you should always be running a primary and seconary dns for that reason as if your dns is down or in your case being updated you won't be able to resolve any thing.

[MickLesk]

What for an issue report is this? No Screenshots? No Output. What failed during Update? My Crystal Ball is currently not working properly.

[MickLesk]

No, simple - thats Not enough. We need clear details. What was the previous version, what happened during the update (screenshots or verbose output), and what isn’t working now? What do you mean by a ‘broken container’? The whole issue is completely unclear. It’s like saying ‘my internet isn’t working’ when I’ve actually forgotten to turn on Wi-Fi. There are some default commands for everyone. Journalctl, systemctl, Logs or an rerun of the Update in verbose.

You cannot assume that we permanently maintain every version of all 580 scripts across all versions

[DFP-]

well, not sure if I should start another issue, but also running into this and I'm seeing the following from technitium.service post-update. May be simple enough and just involving updating .NET

Apr 26 19:49:28 technitiumdns systemd[1]: Stopped technitium.service - Technitium DNS Server.
Apr 26 19:49:28 technitiumdns systemd[1]: Started technitium.service - Technitium DNS Server.
Apr 26 19:49:28 technitiumdns dns-server[2151]: You must install or update .NET to run this application.
Apr 26 19:49:28 technitiumdns dns-server[2151]: App: /opt/technitium/dns/DnsServerApp.dll
Apr 26 19:49:28 technitiumdns dns-server[2151]: Architecture: x64
Apr 26 19:49:28 technitiumdns dns-server[2151]: Framework: 'Microsoft.NETCore.App', version '10.0.0' (x64)
Apr 26 19:49:28 technitiumdns dns-server[2151]: .NET location: /usr/share/dotnet/
Apr 26 19:49:28 technitiumdns dns-server[2151]: The following frameworks were found:
Apr 26 19:49:28 technitiumdns dns-server[2151]:   8.0.22 at [/usr/share/dotnet/shared/Microsoft.NETCore.App]
Apr 26 19:49:28 technitiumdns dns-server[2151]:   9.0.11 at [/usr/share/dotnet/shared/Microsoft.NETCore.App]
Apr 26 19:49:28 technitiumdns dns-server[2151]: Learn more:
Apr 26 19:49:28 technitiumdns dns-server[2151]: https://aka.ms/dotnet/app-launch-failed
Apr 26 19:49:28 technitiumdns dns-server[2151]: To install missing framework, download:
Apr 26 19:49:28 technitiumdns dns-server[2151]: https://aka.ms/dotnet-core-applaunch?framework=Microsoft.NETCore.App&framework_version=10.0.0&arch=x64&rid=linu>
Apr 26 19:49:28 technitiumdns systemd[1]: technitium.service: Main process exited, code=exited, status=150/n/a
Apr 26 19:49:28 technitiumdns systemd[1]: technitium.service: Failed with result 'exit-code'.```

[DFP-]

yep that was it just needed to update aspnetcore-runtime-10.0 via https://learn.microsoft.com/en-us/dotnet/core/install/linux-debian?tabs=dotnet10, dark mode seems to be somewhat glitchy? But I think that's on the app. This logic of checking whether 8/9 are installed and then installing 10 seems to be in the update script? Not sure why it wasn't in my case.

[thecrlgh]

Yes you should always be running a primary and seconary dns for that reason as if your dns is down or in your case being updated you won't be able to resolve any thing.

Wait.......what? With all due respect, this needs to stop being advertised as being updateable, that's misleading then. It's objectively not updateable in its current condition.

if the container as built by the community script lacks a nameserver, then it can't support updates. From an end-user perspective, that's very misleading. I'll get it set to use Cloudflare or something, but I'm surprised to hear that A) it has no nameserver assigned while B) saying it can be updated.

I understand that you can assign a nameserver then run update, but that's not running update, that's an extra step required beforehand.

With all due respect, you may want to take some time and educate yourself a bit more on…well everything.

You didn’t provide any details as to the reason for the failure. You then completely misunderstood why an alternative (backup) DNS is recommended. Finally, you overreacted when the repo’s owner, rightly, informed you he cannot venture to guess the issue based on you providing zero information.

You have an alternative DNS because…things break. It happens - especially with self-hosting AND using graciously-provided scripts we get to use for free.

I updated. Technitium “broke.” It took two seconds to look at the logs, see why, and install the fix. I could do so because I didn’t have a single point of failure (one DNS address).

Even if you weren’t running your own DNS server it’s recommended to have multiple DNS addresses…so if one fails you won’t be “dead in the water.”

Take a breath.