In a total disaster (e.g. fire) recovery situation, with the hardware destroyed, you need to start fresh. A valid approach appears to be : purchase new hardware re-install Linux from distribution (Debian 12 in my case) re-install ISPConfig (3.2 in my case) - with same users and passwords as originally restore config and data from external backups My question is : what exactly should have been backed up externally in the first place ? Some things are fairly obvious : ISPConfig configuration as per 3.1 documentation (/etc, /usr/local/ispconfig, dbispconfig database) /var/www and /var/vmail /var/lib/mysql (takes care of all user's databases) /root/.acme.sh Any additions to /usr/lib/x86_64-linux-gnu/ (e.g. old PHP versions dependencies) Is that all, or am I missing something ? Note: there might be a simpler solution, but short of a physical disk restore, I have not been able to think of one
Run ISPConfig on virtual host, for example install Proxmox VE and create host or several hosts. Then take snapshot backups of the hosts, storing them offsite so fire, flood or earthquake do not destroy both original and backup. Those snapshots of the virtual hosts can be restored on another Proxmox host. I believe they could be restored on other virtualization platforms with a little work.
Or, as an alternative, if you don't want to virtualize the systems yourself, use a reliable cloud hoster. It does not have to be AWS or Google; others provide better value for money for typical hosting setups.
