Answer
Two separate root causes are worth checking on a fresh install that fails this way. First, insufficient CPU/memory on the install VM can make WebLogic server startup take longer than the installer's timeout window, even though the servers often come up fine if checked manually afterward — increasing the VM's CPU/memory allocation resolves this. Second, if the machine originally obtained its network settings via DHCP before being switched to a static IP, leftover DNS-suffix values from that earlier DHCP lease can linger and conflict with the self-signed certificate's required FQDN Common Name during OHS setup — either avoid using an FQDN for the certificate's Common Name, or update the Apache/OHS configuration files to consistently use the FQDN and restart Apache, to resolve the mismatch.