Foreman-installer error going from 1.24.3 -> 2.0.2

bmagistro · June 16, 2022, 2:47pm

When upgrading foreman from 1.24.3 → 2.0.2 an attempt to upgrade postgres occurs but the hooks that check if postgres12 need to be installed/upgraded do not account for all scenarios.

In our case we have a puppet module installed managing postgres (I didn’t think it should have touched tfm but that’s in the past now) which only supports a global version number. At some point postgres was updated to the scl provided version that the installer tries to switch to but the data file was kept in the old location so it thinks it still needs to do the upgrade.

Will be working around this by editing the check locally to return false

github.com

theforeman/foreman-installer/blob/8411f367ce5cb9e185523a42ac3bac54d7379eb9/hooks/boot/01-kafo-hook-extensions.rb#L89

    
      
          
          
def candlepin_enabled?
            katello_enabled?
          end
          
          
def pulpcore_enabled?
            module_enabled?('foreman_proxy_content')
          end
          
          
def needs_postgresql_scl_upgrade?
            !File.exist?('/var/opt/rh/rh-postgresql12/lib/pgsql/data') && File.exist?('/var/lib/pgsql/data')
          end
          
          
def el7?
            facts[:os][:release][:major] == '7' && facts[:os][:family] == 'RedHat'
          end
          
          
def el8?
            facts[:os][:release][:major] == '8' && facts[:os][:family] == 'RedHat'
          end

Running Start Services
================================================================================
Check if command is run as root user:                                 [OK]
--------------------------------------------------------------------------------
Start applicable services:
Starting the following service(s):                                    [FAIL]
No service found matching your parameter 'postgresql'
--------------------------------------------------------------------------------
Scenario [Start Services] failed.

The following steps ended up in failing state:

  [service-start]

Resolve the failed steps and rerun
the command. In case the failures are false positives,
use --whitelist="service-start"

This is here now more for others in case they hit the issue. We are aware these versions are old and are in the process of making our way to current versions.

bmagistro · June 16, 2022, 2:59pm

Minor correction and should have double checked. We did not install the scl version but instead install from the postgres repo which likely accounts for the path mis-match…

bmagistro · June 30, 2022, 7:35pm

I’m going to suggest closing this as our scenario is likely outside the scope of the tfm installer and db manage (I don’t think that’s the correct name) should have been switched off.