elibosley

Please see our post in our public CVE reports board here for updates on this: https://product.unraid.net/p/copy-fail-cve-2026-31431?b=cve-reports

This thread is being discontinued in favor of live updates from Github, follow them here:
https://github.com/unraid/api/releases

We may also post additional changelog updates in the Discord server (https://discord.unraid.net) and will continue to monitor the forums for support, we just won't continue to post release notes here.

It is not but I’m working on a more permanent solution now. These internal keys should be generated on first boot and not regenerated, We will be moving to a more maintainable system in a coming release.

This all seems to be related to the issue we discovered today causing problems with the database backing mothership. Please try signing out and back in and let me know if issues persist.

This was a bug in our registration system. This is now fixed, please try again!

Please update Unraid Connect and let me know if you’re still having this issue!

Can you please install the plugin, and then enter this command:

unraid-api plugins install -b unraid-api-plugin-connect

Then please let me know if that works or errors out and I can help out from there.

We’ve released quite a few updates that fixed docker and lifecycle issues since March, I’d recommend trying out the latest release if you’re willing as I’d love to know if it has solved these issues.

The latest update has fixed this lifecycle issue. Please update your Connect plugin.

This is fixed as of our Monday, July 28 release, we’ve set a timer job to truncate logs over 5mb rather than relying on Logrotate.

This is fixed as of our Monday, July 28th update - we now truncate logs every 20 minutes if they grow over 5 megabytes, which should prevent this going forwards.

Thank you for flagging this. We'll be releasing a hotfix today that should solve these issues.

That prior response is no longer the case. We no longer symlink the API in a way that would break zsh but not bash. I'm not sure why the issue still occurs and will need some community support to figure out what process exactly is failing.


Here is the updated solution that is confirmed working: https://forums.unraid.net/topic/191889-cannot-update-to-the-latest-version-of-the-plugin/#findComment-1568640

I think that the way you're setting up zsh is likely breaking a lot of boot scripts (not just ours).

TL;DR
Yes, your line can and will break Unraid plugin installs by hijacking the login shell. Fix it by guarding with -t 1 to only exec Zsh in actual interactive terminals.


✅ Solution: Detect interactive terminal only
Modify the logic in .bash_profile to only launch Zsh if it's an interactive terminal (i.e., not from a script or system process):
Replace:
echo "if [ -d /mnt/user/appdata/zsh/.oh-my-zsh ]; then if [ -f /bin/zsh ]; then /bin/zsh; exit; fi fi" >> /root/.bash_profile
With something safer:
echo 'if [[ -t 1 && -f /bin/zsh && -d /mnt/user/appdata/zsh/.oh-my-zsh ]]; then exec /bin/zsh; fi' >> /root/.bash_profile
Explanation:
[[ -t 1 ]]: Only true if the shell has a terminal attached (i.e. interactive session)
exec /bin/zsh: Replaces the shell process only in interactive cases
This ensures:
Normal user terminals get Zsh
Plugin scripts, cron jobs, and Unraid boot/install logic remain unaffected


Note: I have NOT tested this logic and using Zsh like this (as default shell) is entirely outside of Unraid supported functionality. You will continue to have issues doing this in the future.

It looks possible that your Connect client isn't installed. Please reinstall the plugin if possible and let me know if you see more options in that menu (flash backup, etc.)

@luve04 Please update your Connect plugin to the latest version. We've just shipped a release that should fix this issue.

I believe I've fixed this now! Please make sure to click on the "Reload" banner in Connect and report back if there are still issues. 

Thanks for reporting - please restart the USB creator and try again if you were having issues, we've fixed the download URL. 

Sorry, we changed plugin URLs today. We've been working with @AndrewZ to minimize this, but something may have slipped through. Don't worry about the "Blacklisted" status, it's temporary and will be lifted once he sets the new URL to be valid. 

Hi!
 
Thanks for bringing our attention to this.
 
Please try again - we've fixed the issue. 
 
Thank you!

Can you give it another check? Should have you sorted out. 

The log max size is set to 10MB, so it should never exceed this. It looks like yours is 7.5M here, so it shouldn't get much bigger before being rotated (aka deleted)
 
We'll be changing this to 2.5MB in the next Connect plugin release.

The log max size is set to 10MB, so it should never exceed this. It looks like yours is 7.5M here, so it shouldn't get much bigger before being rotated (aka deleted)
 
We'll be changing this to 2.5MB in the next Connect plugin release.

It should be fixed now! Just wait for your DNS to clear or try a different browser!

Hi Everyone! We've narrowed this down to a field that used to be inside of the myservers.cfg file but is no longer present. For some reason our ini serializer cannot support this field so it crashes when it tries to write it. 
 
The easiest solution is to delete the file which is located at /boot/config/plugins/dynamix.my.servers/myservers.cfg. This will log you out but My Servers will work after it restarts.