YunoHost 11.0 (Bullseye) release / Sortie de YunoHost 11.0 (Bullseye)

Il y a une branche pour Bulleyes, à tester peut-être ?

(cc @Aleks )

edit: voir le commentaire en dessous :warning:

1 Like

Wasn’t aware of this but I would highly NOT recommend it as it justs uses some custom tools to install Python 2 on the machine and this sounds pretty dangerous. Technically that may work, sure. But that’s a very bad idea.

But Python 2 has reached end of life for more than 2 years now (and it’s been deprecated since at least 5 years) and won’t receive security fixes.

FFsync is also unmaintained since almost 2 years (“recent” commits are basically only about flagging the repo as unmaintained).

So bottom line : don’t use unmaintained software that is built on top of unmaintained technology - period. Yes it’s a shame that Firefox abandoned the project and we could have long discussions about the sustainability / ethic / governance / usability of FOSS but ¯\_(ツ)_/¯

5 Likes

@lapor & @Aleks

Upgraded to ynh 11 without (many) issues, but I have your same issue with gcc-8-basepackage.

Following you advice @Aleks

root@ynh:~# apt-mark showhold
root@ynh:~# apt-mark unhold gcc-8-base
gcc-8-base was already not hold.
root@ynh:~# apt upgrade gcc-8-base
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... 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:
 libgcc1 : Depends: gcc-8-base (= 8.3.0-6+rpi1) but 8.4.0-7+rpi1 is to be installed
E: Broken packages

most important apps in my server work fine (I’ve just to repair Calibreweb) so so far it is not a big issue.

3 Likes

hi!
tried to force upgrade for searx to solve that post-migration python issue, but it didn’t work.
find here the log.

As Searx does not have data stored, you can delete it and install a fresh instance on the same domain/path as before.

2 Likes

I also had an issue with gcc-8-base, so I tried something similar to what is proposed here, and it kind of worked. Kind of, because migration was still failing (something about an unknown option allow_yunohost_upgrade), so I played a bit more (running sudo yunohost tools upgrade system and such). Now I’m in a weird state though… Erreur : La migration 0021_migrate_to_bullseye a échoué avec l’exception La distribution Debian actuelle n’est pas Buster ! : annulation.

$ sudo yunohost --version
yunohost: 
  repo: stable
  version: 11.0.9.6
yunohost-admin: 
  repo: stable
  version: 11.0.9
moulinette: 
  repo: stable
  version: 11.0.9
ssowat: 
  repo: stable
  version: 11.0.9
$ sudo lsb_release -a
No LSB modules are available.
Distributor ID: Raspbian
Description:    Raspbian GNU/Linux 11 (bullseye)
Release:        11
Codename:       bullseye

I think I’m almost there, but how can I wrap up the migration?

Thanks in advance and many thanks to the support team!

1 Like

actually found out that it was already uninstalled somehow?!?
i had to just reinstall it and then it worked fine again.

2 Likes

In fact I found that just had to --skip this migration :slight_smile: after running the rest of the migrations, and a couple of manual tweaks (two missing packages related to php7.4, a locale issue in the migration from postgresql 9 to 11), things are working for me (all apps and web admin are back up – except for ihatemoney but I should be able to fix it later).

1 Like

I am very curious to know what you did to fix the postgresql migration, because that’s the current showstopper for me. Would you mind sharing what you did to overcome or bypass the problem? Many thanks!

So in the logs I found out that there was an issue with something related to the locale. For some reason it expected en_GB.UTF-8 while I only had en_US.UTF-8 (and fr_FR.UTF-8 and a few others). Then I ran sudo dpkg-reconfigure locales, ticked en_GB.UTF-8, and then sudo yunohost tools migrations run. And it worked for me.

3 Likes

Closer investigation of my postgresql migration problem reveals that I have the same error messages as @chmeyer in addition to the locale problem experienced by several, eg. @echapon. The postgresql upgrade logs state

ERROR:  could not access file "$libdir/postgis-2.5": No such file or director
y

And the loadable_libraries.txt file in the log directory says

could not load library "$libdir/postgis-2.5": ERROR:  could not access file "$libdir/postgis-2.5": No such file or directory           
In database: mobilizon                                                                                                                 
could not load library "$libdir/rtpostgis-2.5": ERROR:  could not access file "$libdir/rtpostgis-2.5": No such file or directory       
In database: mobilizon

This is exactly the same as chmeyer’s problem. The postgis package postgresql-11-postgis-2.5 is already automatically installed. Installing postgresql-13-postgis-3 did not help, so I removed it.

The locale problems don’t show up in the postgresql log directory, but in the yunohost migrations log.

After the latest yunohost update yesterday (on 4.3 for version 11.0.9.7) the following error appeared for DOVECOT (mail):

After the obligatory reboot of the server, the dovecot service could not restart because it was not able to find a file anymore (see logs below).

My solution was to locate the missing file ffdhe2048.pem and copy it to the expected location:

  • Expected location: /usr/share/yunohost/other/ffdhe2048.pem
  • Real location: /usr/share/yunohost/ffdhe2048.pem

Afterwards dovecot service was able to start again without problems


  - Aug 18 09:07:26 systemd[1]: Starting Dovecot IMAP/POP3 email server...
  - Aug 18 09:07:26 dovecot[927]: doveconf: Fatal: Error in configuration file /etc/dovecot/dovecot.conf line 26: ssl_dh: Can't open file /usr/share/yunohost/other/ffdhe2048.pem: No such file or directory
  - Aug 18 09:07:26 systemd[1]: dovecot.service: Main process exited, code=exited, status=89/n/a
  - Aug 18 09:07:26 systemd[1]: dovecot.service: Failed with result 'exit-code'.
  - Aug 18 09:07:26 systemd[1]: Failed to start Dovecot IMAP/POP3 email server.
  - Aug 18 09:13:47 systemd[1]: Starting Dovecot IMAP/POP3 email server...
  - Aug 18 09:13:47 dovecot[3897]: doveconf: Fatal: Error in configuration file /etc/dovecot/dovecot.conf line 26: ssl_dh: Can't open file /usr/share/yunohost/other/ffdhe2048.pem: No such file or directory
  - Aug 18 09:13:47 systemd[1]: dovecot.service: Main process exited, code=exited, status=89/n/a
  - Aug 18 09:13:47 systemd[1]: dovecot.service: Failed with result 'exit-code'.
  - Aug 18 09:13:47 systemd[1]: Failed to start Dovecot IMAP/POP3 email server.

The good solution is to force-regen the conf for dovecot with yunohost tools regen-conf dovecot --dry-run --with-diff and yunohost tools regen-conf dovecot --force

4 Likes

Thank you. I’ll do this later today

1 Like

@YouKnowHorst I run into the same issue !

When I read the line 26 of file /etc/dovecot/dovecot.conf on line 25 in comment there is the curl command to fetch the last certificate such as below

curl https://ssl-config.mozilla.org/ffdhe2048.txt > /path/to/dhparam

I did that command with just correct path to dhparam and voila :upside_down_face:

Yes, technically that works, yet that’s not the proper fix. If everybody had this issue, we would have taken care of it, so maybe ask yourself 2 minutes why you have this issue and not the others. The reason is probably that you manually edited dovecot’s conf file therefore the regen-conf is not upgrading it, and in particular it doesn’t upgrade the line mapping to the certificate.

So what you should do is run yunohost tools regen-conf dovecot --dry-run --with-diff, review the diff, and yunohost tools regen-conf dovecot --force if you’re happy with the diff.

4 Likes

I did modify dovecot conf already indeed :face_with_open_eyes_and_hand_over_mouth:

Everything went well except synapse does not start, here’s the log: https://pastebin.com/dGtvh6hL

The app-upgrade

yunohost app upgrade -f synapse

gets me

Error: All apps are already up-to-date

yunohost service restart matrix-synapse

gets me

Aug 18 16:39:49 python[5422]: /opt/yunohost/matrix-synapse/bin/python: Error while finding module specification for 'synapse.app.homeserver' (ModuleNotFoundError: No module named 'synapse')

Port 8448 is not reachable - which was okay before the Bullseye-upgrade. I’m on a Hetzner VPS without firewall on that VPS.

Anyone any idea? Thanks!

unfortunately -f is not the appropriate option, you may retry with either -F or --force

2 Likes

Aha! That’s it! Yes I’ve modified the dovecot config as well to allow more connections :slight_smile: I did totally forget this!