Open alexAubin opened 2 years ago
Look for old legacy stuff in the core that can be dropped out
I can find:
During the postinstall I get:
Warning: ls: cannot access '/etc/php//fpm/pool.d/.conf': No such file or directory Warning: adduser: Warning: The home dir /var/vmail you specified can't be accessed: No such file or directory Warning: grep: /run/resolvconf/resolv.conf: No such file or directory
Here the full logs, but from what I understood:
After the postinstall, even though YunoHost-API was started, I couldn't access the webadmin (error 502), I had to restart it
Error from nginx:
==> /var/log/nginx/bookworm.tld-error.log <== 2023/09/11 09:22:25 [error] 3616#3616: *6 connect() failed (111: Connection refused) while connecting to upstream, client: 10.34.87.1, server: bookworm.tld, request: "GET /yunohost/api/installed?locale=fr HTTP/2.0", upstream: "http://127.0.0.1:6787/installed?locale=fr", host: "bookworm.tld", referrer: "https://bookworm.tld/yunohost/admin/"
During nextcloud install, fail2ban took too much time to reload (you can find some warnings). The issue here is the line_match
arg we should match (Started|Reloaded) fail2ban.service - Fail2Ban Service
(there is a fail2ban.service
to add)
There is also a warning in fail2ban service:
fail2ban.configreader [33783]: WARNING 'allowipv6' not defined in 'Definition'. Using default one: 'auto'
Moreover, I tried to get banned by entering the wrong password several times, but I didn't succeed.
That's all I've been able to see so far after installation
Things I have noticed while upgrading non YUNoHost systems from bullseye to bookworm that could have an impact on YUNoHost :
running postinstall on a fresh bookworm rebuild incus container:
postinstall fails because there is no dns resolvconf: https://paste.yunohost.org/nukegaqoca.sql
root@ynh-dev-bookworm-unstable:/ynh-dev# cat /etc/resolv.conf
# Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
# DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
# 127.0.0.53 is the systemd-resolved stub resolver.
# run "resolvectl status" to see details about the actual nameservers.
nameserver 127.0.0.1
root@ynh-dev-bookworm-unstable:/ynh-dev# resolvectl status
bash: resolvectl: command not found
root@ynh-dev-bookworm-unstable:/ynh-dev# systemctl status systemd-resolved
Unit systemd-resolved.service could not be found.
I think systemd-resolved
should be installed, and I don't understand why it isn't.
Trying to fix it:
root@ynh-dev-bookworm-unstable:/ynh-dev# echo "nameserver 1.1.1.1" > /etc/resolv.conf
root@ynh-dev-bookworm-unstable:/ynh-dev# apt update
Hit:1 http://forge.yunohost.org/debian bookworm InRelease
Hit:2 http://deb.debian.org/debian bookworm InRelease
Hit:3 http://deb.debian.org/debian bookworm-updates InRelease
Hit:4 http://deb.debian.org/debian-security bookworm-security InRelease
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
All packages are up to date.
root@ynh-dev-bookworm-unstable:/ynh-dev# apt install systemd-resolved
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:
The following packages have unmet dependencies:
yunohost : Conflicts: systemd-resolved but 252.22-1~deb12u1 is to be installed
E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.
Edit: no systemd-resolved
shouldn't be installed. Then, I don't know what to do yet
After the postinstall, I tried to connect to the webadmin, pressed the Connect
button, nothing happens.
URL: https://ynh-dev-bookworm-unstable.incus/yunohost/admin/#/login?redirect=%2F
I changed URL to https://ynh-dev-bookworm-unstable.incus/yunohost/admin/, I'm logged in, works as expected.
My guess is there is #/login?redirect=%2F
added at the end of the url which shouldn't be there, or we should redirect #/login?redirect=%2F
to #/
when use is logged in
When I CTRL + SHIFT + R
(refresh without cache), there is a [GET] /domains
in the history, at the bottom of the webadmin (it's green, but it's always here, idk if it's normal)
Remaining major items :
I'd say the apps isn't a blocker anymore: 49 apps is less that 10%.
Tested on my side with ~30 apps.
Works well, but I have to change the ssowat conf from basic-without-password
to basic-with-password
for some apps (roundcube and aeneria).
Moreover, I had to add "protect_against_basic_auth_spoofing": false
to make nextcloud works on my PC and my phone.
And the python venv migration fails for all apps (which should be fixed by https://github.com/YunoHost/yunohost/pull/1912, but I can't test it after the migration), but a simple upgrade of the corresponding apps fixed the issue.
Same as kay0u. Tested with 42 apps. Here are the issues I had :
:warning: Thread policy : following past traumas, we will not hesitate to remove any comment which do not add anything constructive on making progress with this, such as asking "Any update on this ?", "What is the ETA ?", "Bookworm is soon to be released", "Bookworm has been release for X days zomg 1!?!!", etc. We are a volunteer team, we will release stuff when it's stable, and you are welcome to effectively help us getting shit done, but for the love of god don't pressure volunteers as if they owe you something
Relevant links
Core TODO
Major stuff to change in YunoHost ?
aptitude
instead ofapt
, at least for the migration procedure, to avoid all the apt/dpkg hell being absolutly stupid at resolving complex situations ...Bullseye -> Bookworm support
Implement the php migration (in the bookworm branch)Apps
Relevant links:
Difference between bulleyes and bookworm CI results for applications
[x] Start an app testing campaign to identify regressions
[x] Tweak dash.yunohost.org / Tartiflette to display this on a nice dashboard
[ ] Fix regressions found in apps