How to troubleshoot a failed Debian Buster to Bullseye upgrade

Discussion in 'ISPConfig 3 Priority Support' started by webguyz, Jul 30, 2024.

  1. webguyz

    webguyz Active Member HowtoForge Supporter

    Hi,
    Have a multiserver ISPConfig system and have done a lot of slave upgrades from Buster to Bullseye without issue. My Master is still running Buster just fine with no problems so last night I attempted to do the upgrade to Bullseye. Upgrade seemed to go OK using the same 2 HowToForge docs I successfully used for other previous slave upgrades. After I rebooted I got a connection refused instead of the web console I was expecting. I panicked and wasn't sure what to start checking first. Rebooted again. I never got to the part in the instructions where you change the default web settings to 7.4. I was so freaked out I just downed the server and restored my Master from a backup. Was afraid of having the Master down for long but primarily just not sure where to start looking. Obviously web related but I thought maybe the Bullseye switch from using 5.6 to 7.4 PHP somehow screwed up something.

    Could I get some tips on where to start looking in a situation like this. Was not expecting it to not work :) Is there a log I could have looked at generated by the upgrade and where might it be located. I guess the Master can stay offline as long as it wants but no updates or Control Panel changes can be made. I know enough about Linux to be dangerous but when the mess hits the fan I crumble until I get some direction.

    Thank all.
     
  2. Taleman

    Taleman Well-Known Member HowtoForge Supporter

    Check system is running,
    Code:
    systemctl --state=failed
    Read syslog.
    Don't Panic!
    Debian 10 Buster should have PHP 7.3.
    If you suspect the Debian distribution upgrade went wrong, read the Debian Release notes for Bullseye, https://www.debian.org/releases/bullseye/releasenotes, it has info on how to do the upgrade and how to solve issues with the upgrade. Also read again the upgrade instructions you used, maybe you missed some point.
    If you run a virtual machine, make a clone and test upgrade on that clone copy in test environment where it can not connect to the other hosts in your ISPconfig cluster.
     
  3. webguyz

    webguyz Active Member HowtoForge Supporter

    Taleman,
    I do have a virtual server setup and should have set a Checkpoint before starting, but got careless since other upgrades always worked fine. Need to get in the habit of reading release notes. Thanks for troubleshooting suggestions. Will attempt again this weekend.
     
  4. till

    till Super Moderator Staff Member ISPConfig Developer

    And one more thing in addition to what @Taleman posted already: The ISPConfig GUI is basically just a website. If you can't access it, check the global Apache or Nginx error.log file (/var/log/apache2/ or /var/log/nginx/error.log) for errors. Also, an ISPConfig update with 'reconfigure services = yes' after doing the system upgrade often helps. If you can't do that because your system already uses the latest version, use:

    ispconfig_update.sh --force

    command, to override the version check.
     
    webguyz likes this.

Share This Page