On 19/10/2024 07:32, Marc Haber wrote: > Hi, > > during a restore operation, I have recreated a RAID array using a rescue > system. That new array ended up being named grml:0. Renaming the array > to realhostname:md_root didn't work inside the rescue system > > The rescue system is reasonably current, using kernel 6.11.2 and mdadm > 4.3. > > First, I tried changing the name of the array. That worked: > # mdadm --assemble md_root --update=name --name=md_root /dev/sda3 /dev/sdb3 > mdadm: /dev/md/md_root has been started with 2 drives. > # mdadm --examine /dev/sda3 > ... Name: grml:md_root (local to host grml) ... > > What didn't work was changing the homehost: > # mdadm --assemble md_root --update=homehost --name=realhostname /dev/sda3 /dev/sdb3 > mdadm: /dev/sda3 has wrong name. > mdadm: /dev/sdb3 has wrong name. I think instead of --name=realhostname you need --homehost=realhostname here. > > Adding --force didn't work, same error message. > > What is the recommended way to create an array inside a rescue system > that has homehost and name to the values the local admin wants, and/or > to change those values for an array that has already been created and is > filled with data? > > Greetings > Marc > Geoff Back What if we're all just characters in someone's nightmares?
