ncoker If you have to to this: rpm -e dracut-kiwi-lib dracut-kiwi-live dracut-kiwi-oem-dump dracut-kiwi-oem-repart kiwi-systemdeps-bootloaders kiwi-systemdeps-core kiwi-systemdeps-disk-images kiwi-systemdeps-filesystems kiwi-systemdeps-iso-media kiwi-tools you have a very old installation. Yes, these packages are unused and needs to be removed.

    WalterH If 2022.5.1 is very old, then yes. The packages came from the original applicance installation last year. I had to erase it too.

    ncoker Here are our steps summary:

    Thanks for your steps. After

    ncoker zypper --releasever=15.4 dup --download-in-heaps

    there is the error again, I've described in my post

    I hope for any hints. At the moment I try to upgrade my test vm.

    Thanks
    Ulf

    Im also very interested in this and wanted to ask this now.

    It seems there are some strange problems since Christmas-Update. I also use the Appliance, since 2021 and always kept it up to date so far.

    WalterH:
    I also encountered this, but dont understand why it seems to be a problem to have an "old" installation like me. I always installed updates and everything. So it seems to be more of a problem of the update process?

    I installed Christmas update but it always still says I have 2021.05.2 as said here. There seems to be something strange and I have the feeling I dont got all updates since then. For example: I still have the problem I described here, its not fixed for me.

    So I dont know if this is all about wrong dependencies or the lack of the distribution upgrade until now - but would be happy if there is a solution soon.

    Have you tried to setup a fresh ISO-Installation and just used your Backup?
    I'm not done yet with all my Upgrade-Tests to 15.4 but i think this will just be my "Last-Resort" and it is always good to check if that is working anyhow 😊.

    • tf91 replied to this.

      crpb
      I havent tried this yet, because its much more effort and after the messages here that there will be a solution and instructions, I didnt thought that would be nessecary to do a whole New Installation.

      Also there are reports that the ISO is broken/buggy too.

      I just did an ISO-Installation for tests this morning.
      Only thing i had to do was an zypper modifyrepo --disable repo-backports-update and zypper in -y open-vm-tools so far. However i only use core and only test something else if i'm really bored....

      I also opened an ticket because the OVA-Image does not align with the configured Repositorys in the ISO. Meaning that the OVA-Image sadly still uses a few 15.3 Repositorys.

      For this "2021.05.2" the Package grommunio-release-$NUMBER is responsible to my knowledge.
      What are your current Package-Versions?
      rpm -qa *grom*

      • tf91 replied to this.

        Is it not possible to choose another way than zypper dup?

        I have experienced that for a first installation it is best to use only the ISO and not the OVA, without any update during their installation. What if the same method is used for an existing installation, installing the v2022.12.1 over the v2022.05.x?

        Or would there be a way to do a fresh install for as major upgrade and then reliably migrate all user data from install to install, it means from v2022.05.x to v2022.12.1?

        What I read here about zypper dup seems not process safe (!)

        • crpb replied to this.

          Andy Or would there be a way to do a fresh install for as major upgrade and then reliably migrate all user data from install to install, it means from v2022.05.x to v2022.12.1?

          The reliable way is your Backup and Recovery strategy 😛.
          It isn't that much what you need to do.

          My and others Notes about it here which i've basically done now a couple times..
          https://community.grommunio.com/d/307-mailbox-backup/5

          zypper dup is the way of doing an System-Upgrade in Suse. No way around it AFAIK..

          OK. On the linked page I have seen the Restore way and what is the process for the Backup?

          My way of backing up is stated directly above.

          Sorry, I can't read anything from it, I'm not enough of an expert, so it's all unsure for me. Something like a summary is missing. It would be good if there was a kind of script of function for "normal users", or even better, a function in the admin "Save" and "Restore".

          Backing up two directories completely and restoring them completely on the new installation is not the issue.

          /etc/gromox/
          /var/lib/gromox/

          Then backup and restore the databases. I can't do anything with the shown restore sequence at "MYSQL-DB zurückspielen". Could I use it 1:1? Etc. and so on. You can see that these operations cause headaches for the not so experienced users.

          • crpb replied to this.

            I still have "Write a decent Backup-And-Restore-Utility in XYZ" on my ToDo but hadn't had the time for it ☹️.

            You should definitely read this to get a Basic understanding and maybe think about how you could integrate an Backup which fits your needs.
            https://docs.grommunio.com/admin/operations.html#backup-disaster-recovery

            A VM-Snapshot can't be considered as an real backup.
            It doesn't matter about which Software we are talking.

            Just think about when you only want to restore Mailbox A but not B-Z...
            Yes there is stuff like Mailstore/Benno/Whateverelse which i also use but this isn't working for e.g. Contacts or the Calendar.
            For those i got an Script from another Community-User here which i haven't even came around to publish or integrate in my Systems. Notes and Tasks aren't covered with these either.

            Andy I can't do anything with the shown restore sequence at "MYSQL-DB zurückspielen".

            Take a look into the File /etc/gromox/mysql_adaptor.cfg

            eval `sed 's/ //g' /etc/gromox/mysql_adaptor.cfg`

            and after the Command you can do something like echo $mysql_username in the Shell.

            But you can do those just manually like the Link states.
            The eval-Command will basically do the same as if you would have done by cat /etc/gromox/mysql_adaptor.cfg and copy/pasted this output in the same console. Note that eval can be considered "BAD" in many cases. But there is enough about that in the Internet.
            Because i know which content it has i do that w/o worrying about it in any way.
            And that's enough for now.

            I run production and test VM's on Synology NAS. For backup & disaster recovery, I have the option there to export the shutdown VM and import it if necessary. For intermediate steps before updates I create snapshots. I use both of these on a regular basis.

            However, when reading the contents of the linked Backup & Disaster Recovery website, I have doubts that such a backup is suitable for simply restoring after a major upgrade. What I read there rather leaves the impression to create a backup, which can be easily restored in a version clean installation, so I could do that with the above mentioned exports and snapshots. But our topic is something different because of major version jumps.

            For example, if I have a directory in an installation that contains the complete setting and user data and only that, it is very easy to bridge a major version jump, because I copy the directory into the new installation and it runs as before. If the MariaDB would be the database that controls and maps everything, it would be a further simplification, because then the content could be transferred easily via a transfer script.

            Now then, these are thought experiments for a simple handling of such upgrades. From the current point of view, I would not consider any of the given ways as applicable and safe. Maybe the easiest way could be to transfer user data between two running installations via a command sequence.

            crpb
            My current package versions:
            grommunio:~ # rpm -qa *grom*
            system-user-gromox-2-lp153.13.1.noarch
            grommunio-setup-1.0.68.bb1014d-lp153.7.1.noarch
            gromox-debuginfo-2.1.3.cb51757-lp153.54.1.x86_64
            grommunio-antispam-3.4-lp153.2.1.x86_64
            grommunio-admin-api-1.9.26.d9367fd-lp153.45.1.noarch
            grommunio-common-10.e94d08a-lp153.7.1.x86_64
            grommunio-release-2022.05.2-lp153.2.1.x86_64
            grommunio-cui-1.0.239.ad62461-lp153.44.1.noarch
            grommunio-chat-6.2.2-lp153.2.1.x86_64
            grommunio-index-0.1.16.e01e06c-lp153.23.1.x86_64
            grommunio-web-3.1.182.52d9180-lp153.139.1.noarch
            grommunio-jicofo-2.0.5390-lp153.12.4.noarch
            grommunio-imapsync-2.200-lp153.4.1.noarch
            systemd-presets-branding-grommunio-2022.12-lp153.1.1.noarch
            system-user-grommunio-2-lp153.11.1.noarch
            grommunio-keycloak-20.0.2-lp153.18.1.noarch
            grommunio-auth-21.c8473e9-lp153.12.1.noarch
            gromox-2.1.3.cb51757-lp153.54.1.x86_64
            grommunio-sync-1.1.55-lp153.61.1.noarch
            grommunio-files-10.9.1-lp153.17.1.noarch
            grommunio-error-pages-1.0.6.9c50afb-lp153.7.1.noarch
            grommunio-admin-web-2.6.0.33.45689ef-lp153.31.1.noarch
            grommunio-dav-1.1.33.a126414-lp153.1.1.noarch
            plymouth-theme-grommunio-1-lp153.11.4.noarch
            gromox-debugsource-2.1.3.cb51757-lp153.54.1.x86_64
            grub2-theme-grommunio-1-lp153.11.4.noarch
            grommunio-videobridge-2.1.508-lp153.16.4.noarch
            grommunio-admin-common-6.cb985db-lp153.14.1.noarch

            I tried to "upgrade" (= zypper dup) my GRO. Which was a bad idea. It totally messed up my GRO. System had problems each and every day. I'd to do a fresh install. Now I only use the "zypper ref && zypper up" command. Still kind of curious whether someone already upgraded the GRO successfully?

            Is there any other command, skript or readme on how to upgrade a GRO properly? I will need to upgrade my GRO when the new Q1/2023 release with full ews support for macos is out.

            Will I have to do a full, manual (new) installtion and then restore the email accounts and data (I saved them via outlook as pst-files and later restored them)?

            Or ist there a way to "just" do a full new installation in a 2nd VM and then use a skript to transfer settings (i.e. main.cf, master.cf, sasl_password, etc.) email accounts and data via a skript?

              In a fresh installation there is the same dependency problem. This lines come from a fresh appliance installation with core, meet and files (grommunio-setup.log).

              Paketabhängigkeiten werden aufgelöst...
              
              Problem: das zu installierende grommunio-files-10.9.1-lp154.1.1.noarch erfordert 'php7-smbclient', aber diese Anforderung kann nicht bereitgestellt werden
                Nicht installierbare Anbieter: php7-smbclient-1.0.0-bp154.1.67.x86_64[base]
               Lösung 1: Folgende Aktionen werden ausgeführt:
                Deinstallation von php8-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-zlib-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-zip-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-xmlwriter-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-xmlreader-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-sysvshm-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-sqlite-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-sodium-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-soap-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-posix-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-pdo-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-pcntl-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-openssl-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-opcache-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-mysql-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-mbstring-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-iconv-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-gettext-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-gd-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-fpm-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-dom-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-curl-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-ctype-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-cli-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-bcmath-8.0.27-150400.4.23.1.x86_64
                Deinstallation von php8-redis-5.3.7-lp154.1.1.x86_64
                Deinstallation von grommunio-sync-1.1.62-lp154.75.1.noarch
                Deinstallation von grommunio-web-3.1.197.c7957cc-lp154.157.1.noarch
                Deinstallation von grommunio-dav-1.1.37.d5b6463-lp154.10.1.noarch
                Deinstallation von gromox-2.2.140.7f3e6a0-lp154.112.1.x86_64
                Deinstallation von patterns-grommunio-1-lp154.8.1.x86_64
               Lösung 2: grommunio-files-10.9.1-lp154.1.1.noarch nicht installieren
               Lösung 3: grommunio-files-10.9.1-lp154.1.1.noarch durch Ignorieren einiger Abhängigkeiten brechen
              
              Wählen Sie aus den obigen Lösungen mittels Nummer oder brechen Sie (a)b [1/2/3/a/d/?] (a): a
              2023-02-03 11:18:06 ::  Dialog: mysql installation type
              2023-02-03 11:18:12 ::  Selected MySQL installation type: 1

              Perhaps on weekend I have the time to test a restore.

              4 days later

              In the meantime I learned that the files-plugin is still based on php7. So I will wait until all modules are converted to php8 before updating.

              • Andy replied to this.
                7 days later

                Just did an upgrade at our internal System.
                Everything seems to be in order expect a few errors in nginx-web-error.log about fastcgi but i couldn't find anything not working so i will wait for the Support what that might be..

                DAV works again!++++

                vdirsyncer works

                  BusinessTux all modules

                  I have no idea where are the modules to configure. Is there a special modules folder?

                  © 2020-2024 grommunio GmbH. All rights reserved. | https://grommunio.com | Data Protection | Legal notice