We have created a new test server on Ubuntu 22 and am migrating some sites over from our old server which was Ubuntu 18. I have tried restoring from backups already created and stored, creating new backups with different settings that leave out the SSL settings altogether but no matter what the restore fails because it cannot find the same SSL that it was sharing on the original server.
Can you advise what options I need or need to omit in order to get the server restore to work please ?
This doesn’t make full sense, have requsted a new certificate using let’s encrypt ? You put question marks in the system details for web server (the chioce here would be either apache or nginx) this may have a baring on the issue. Are you using cloudflare?
Oh I see thankyou for the clarification - its Apache we are using.
Basically we do not get to an opportunity to do a certificate with Lets Encrypt because the restore doesn’t continue past the message. When doing the restore, we click Restore all options then deselect the Apache SSL configuration and certificate and then hit proceed. The summary of what it will install then shows
" Apache webserver configuration
Mail/FTP users and mail aliases
Logrotate configuration for log file
Records in DNS domain, except SOA
Contents of server’s MariaDB databases
Virtual server password, description and other details
AWStats configuration file"
Then as it proceeded with the extraction and install it would stop on the error message about it could not find the shared SSL certificate (which was on the original server).
what does this mean ? most domains use their own certificate, which is not normally shared.
if you view the archive you are going to restore do you see the directory .backup , if so can you see the certificates there <domain-name>_ssl_<file> where <file> is cert, key and ca
I could not agree more - hence the post! It was shared because it was a subdomain to a parent which had the certificate, so on the original I could see what it meant, what didnt make sense was the fact that it was whining about it but I was not trying to move or copy the SSL details.
In trying to run it so I could copy some details for you direct from the server in question, it has just restored the virtual without issue - using the same settings I previously used and described in the post. I cannot confirm whether this was a server or a virtualmin bug or abberation, only that we were meticulous with the process and actually went through it repeatedly as a team to make sure.
I am not sure what to take from this in going forward but thanks for your reply and attempts to help jim
Ah I see, as a footnote I always request a certificate for a sub server in it’s own right whether it is a sub domain of the parent or not, hence never seeing the problem
I found the template the system was made with server template 14 years old does act the same way that a new default server template acts. And can you change it but, the new one makes links to Vertualmin,postfix, dovecot; the old one didn’t and they are not created.
Just something to watch for in the SSL upgrade I found…
