Re: best way to replace all disks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



unfortunately I can't change the server. I have tons of custom,
closed-source software
that won't run on newer hardware, that's why i have to use the current one.

Yes, i'm planning a replacement, but is not something quick to do,
i'll need many month
redeveloping, in-house, the current closed-source software. Until
that, I have to keep the
data safe.

technically I can use a new hardware, but I can't upgrade the
operating system (a very old slackware,
if I remember properly) or these closed-source software wont run anymore

maybe the faster solution would be to add a cheap (which one?) HBA
controller that supports more than 4 SATA disks,
move all disks to this new controller and then convert, one by one,
each 2way mirror to a 3way mirror:

mirror01: sda, sdb will be converted to sda,sdb,sdf, then replace sda
with sdg, then remove sdb downgrading to a 2way mirror
but with all new disk. During the whole operation, redudancy is never
lost and I'll do this on a live disc, with the original raid unmounted

Il giorno lun 22 apr 2019 alle ore 16:24 Carsten Aulbert
<carsten.aulbert@xxxxxxxxxx> ha scritto:
> Frankly, I would opt either to get a new server - depending how old very
> very very very old is - or boot a live disk and copy the *contents* of
> everything over to one of the new disks attached to the system via USB
> or any other means (spare on-board SATA) as my guess would be a single
> disk should be able to hold everything you currently have.
>
> After this was done successfully, put all old disks out of the system as
> a safe-guard if something goes terribly wrong.
>
> Then put in all new disks EXCEPT the one where your copy is on. Boot the
> live system again, create your mirrors/raids (some may be incomplete at
> the start) as you like and then start copying data back from the "copy
> disk".
>
> Once done and verified[1] you should be able to re-partition the "copy
> disk" and make all mirrors/raids complete again.
>
> Having all that said, for servers older than 8-10yrs it would probably
> be better to get a new set of hardware and copy data between both systems.
>
> Cheers
>
> Carsten
>
> [1] if you OS is on these disks as well, you may want to check if you
> can boot your system from the new disks!
> --
> Dr. Carsten Aulbert, Max Planck Institute for Gravitational Physics,
> Callinstraße 38, 30167 Hannover, Germany
> Phone: +49 511 762 17185




[Index of Archives]     [Linux RAID Wiki]     [ATA RAID]     [Linux SCSI Target Infrastructure]     [Linux Block]     [Linux IDE]     [Linux SCSI]     [Linux Hams]     [Device Mapper]     [Device Mapper Cryptographics]     [Kernel]     [Linux Admin]     [Linux Net]     [GFS]     [RPM]     [git]     [Yosemite Forum]


  Powered by Linux