I have searched the existing open and closed issues
Required Information
DietPi version | 9.10.0
Kernel version | 6.6.74
SBC model | RP4B
SD card used | USB SSD
Additional Information (if applicable)
Software title | (PiHole & Nextcloud)
Was the software title installed freshly or updated/migrated? → Updated
Steps to reproduce
I updated via pihole -up
Expected behaviour
expected everything to work
Actual behaviour
phone was giving me error messages it can not sync to webdav calendar anymore. Checked the webinterface for nextcloud and can not access ist anymore (internet works fine). Pihole webinterface works fine, too.
Extra details
dietpi-service status shows lighttpd.service failed (reinstall of lighttpd did not solve the issue)
Just did a complete reinstall of the whole system, nextcloud still can’t be accessed.
I switched to Beta Channel 9.11. and did a complete reinstall of Dietpi. I am now running PiHole 6 and Baikal for CalDav/CardDav. Works fine. Will skip nextcloud since I only ever used the dav funcionality.
Got the same problem here. PiHole wasn`t asking anything while updating. But as Joulinar suspects: It has to do be a part of the webserver. When i open my Nextcloud-Bookmark suddenly PiHole pos up…
Since i have large ammounts of data on Nextcloud, and i dont want to roll back with the backup, does anybody have a clue how to make nextcloud answer its ip? nextcloud-adress/nextcloud/
I encountered what people described in the current thread but I don’t understand what the fix is. Should I restore a backup and disable pihole updater? How to do this effectively? I know there is a cron script but I’m not sure if this is the only pihole self-update method.
Edit: OK! The solution was just to update to 9.11 beta.
yes, you can upgrade to the latest DietPi v9.11 BETA, which we will release today or tomorrow at the latest. This will fix systems that have already migrated to PiHole v6, as we will move PiHole v6 to port 8089 to avoid port conflicts with running web servers or other applications.
I’m a longtime pihole user and if there’s anything I can pass on, it’s to always wait at least 48hours before updating . At least they are quick to fix things
yes, our developer @MichaIng has added some fixes as well. Hope it will become stable soon. At least my test migration was working without issues today and all information got transferred/migrated.
Yeah, generally good to give it 48h, I agree, and this one is a major version increment with two non-trivial migrations:
Merging all the dedicated config files for Pi-hole and dnsmasq into one.
Migrating from Lighttpd to internal webserver + PHP interpreter.
Here, a week + backup or testing on clone before updating production makes much sense.
However, since it required us to change DietPi as well, I followed and worked on remaining issues closely this weekend, and indeed the major issues have been solved.
For DietPi users, as long as you do not have Lighttpd as webserver installed, the Pi-hole v6 update won’t offer to disable any other webserver, and not PHP in any case. Hence the DietPi update as follow-up makes sense, to clean-up things. We do not want to force anyone to update to Pi-hole v6, hence the DietPi-specific cleanups will be offered on every next DietPi update for Pi-hole instances updated in the meantime, until we are certain there are no Pi-hole v5 instances anymore, or that we can safely enforce the update to v6. If the webserver is not disabled, Pi-hole will run on ports 8080/8443 after the update to v6. The DietPi update will however switches it to 8089 only, and no HTTPS, to avoid any conflicts with other web applications. The update also informs about that.
. At least they are quick to fix things