Hello Guys



I need some help because I will have to replace a disk in a Linux Workstation and rebuild its Software Raid1. For Sure before start I will run a full data backup but I never did a RAID repair, so I need to be sure about the step I have to follow. The system should be configured like below, could you tell me if the procedure i wrote is correct or no:







I have sda2 that is marked as failed I have to replace I need to mark all the partition /dev/sdb as failed



m



then I will remove all the partition from the raid:





Then I shut down the system and I will replace the disk where I have to create the same partition type I have in sda







and then I will add the partition to the raid system







Is it correct what I wrote? I hope yes, now what it make me more concerned if the RAID software needs boot in UEFI mode. If I remember well in this case I should have a standard Partition /boot/efi and I hould have a situation like this



/

Now If the broken disk were sda we need to repair not only the Raid partitions but also reinstall the /dev/sda1 with the efi partition to make the system bootable again. In this case there is a way to proceed?





Thanks in Advance.

Gio

You should install a copy of the bootloader and efi partition on drive sdb so that if sda fails you can remove it and boot from sdb. The copy will not be managed by RAID, but it seldom changes so you can manage it yourself.

In your description, you say sda2 is failed. Do you mean sdb2? Make sure you know which drive has failed by issuing read commands to the drive. Also make very sure you know which physical drive it is in the system by observing the disk select lights. Your steps are correct if sdb is the failed drive.

Presumes the system will boot with the RAID in degraded mode. The initrd in most distros should handle it ok, but I recall Debian not having the logic in there for quite a while - hopefully they also have it sorted by now.
In need you could always use a liveUSB.