This method is only valid for systems with CentOS 7.4, an only when it was installed in UEFI mode
Note: UEFI is the default Mistika iso installation mode for Z8 systems since March 2018 ( that is if the computer support UEFI and it does not have the BIOS legacy mode activated, otherwise the BIOS legacy mode is used )
Please note that this article is not for cases when the system start booting normally but it does not reach the graphics mode. This article is only for the case that the system does not even reach the grub stage (which is the first boot menu offering the CentOS Linux boot), and it is only for the case of UEFI boot installations, not for boot installations using the BIOS legacy mode). Please open a support case if you are not sure about this.
Note: SGO only supports UEFI mode for Hp workstations models from 2017 and later (Z8, Z4, etc). Older systems (Z820, Z840 ... ) should be always installed in BIOS legacy mode, as they are known to experience problems with UEFI modes (even losing the partition table). Please ask support if not sure.
In UEFI systems, the "no operating system" boot error usually means that the centos EFI boot entry in the NVRAM of the computer (not in the hard disk, but in the non volatile memory used by the firmware) has been lost or damaged in someway, or that it is not valid anymore due to hardware changes (for example when alternating between different boot disks). Many systems can autodetect this situations and find EFI boot files correctly, but some computer models and particular CenOS installations may fail to find the correct boot file automatically, so it may need to be reinstalled in the NVRAM again. This procedure will also install it in the default location, which make it more resilient agaisnt future incidents.
Obviously the other possibility is if the boot disk itself (or its EFI partition) has been damaged rather than the NVRAM content, but if you are not sure it is worth to try this.
In a situation like this you can try this method to repair the UEFI boot manager entry:
1 - Disconnect all storage except the boot disk (disconnect disk arrays, internal storage, external disks and pen drives, etc. Only the boot disk should be still there!)
Try to boot in that simple configuration. This is to discard problems related with the presence of other storage devices.
If still does not work:
2 - Now boot from a pen drive with the Mistika installation ISO. Ask email@example.com if you do not have it ( or use any other CentOS iso or live pen drive that you could get from other sources, as it is not for reinstalling but to use it for this fix ), and select the "Rescue linux installation" menu.
3 - This will try to find the boot disk, and it will mount it under /mnt/sysimage. If it success it will offer to mount it.( Press "1" to continue), and follow the next steps (if it fails please contact support)
- Now execute:
This will tell you the EFI boot partition for your hard disk, which is the partition appearing in the line ending in /boot/efi . For this example, let's suppose that it is /dev/sda1 (do not continue if you are not sure about the exact partition id)
4 - Now execute this to recreate the boot manager entry for it:
efibootmgr --create --label CentOS --disk /dev/sda1 --loader "\EFI\centos\grubx64.efi"
cp /boot/efi/EFI/centos/grubx64.efi /boot/efi/EFI/BOOT/grubx64.efi
And execute this to exit and reboot:
That will reboot the system. If it reaches the linux grub menu then you have succeed. Now power off the computer and reconnect all the storage that was disconnected at step1.
5 - Good luck...