Re: RAID & Backups
On Thu, Sep 04, 2025 at 12:01:20PM +1000, Geoff wrote:
> One point is that you have to be actively monitoring for any drive failures.
> Otherwise after one fails the server will still be working and will just be
> sitting in a rack in a DC with a flashing red light that no one will notice
> until another drive fails and service is lost.
I have had drives fail and received much email from the system whingeing about
it. This is with Linux machines using MD as RAID (ie Linux kernel/software
RAID). Your alias for root in /etc/aliases needs to point to you.
RAID gives you time to fix/replace the failed drive. That time < forever.
When you set up a new RAID-1 system you have to buy drives of the same size.
Avoid the temptation to buy two of the same disk, ie same model/manufacturer.
They will prolly be from the same batch if you buy them together and so may
have the same failure characteristics. You use them in an identical way (that
is what RAID-1 means), they are in the same machine (so same temperature, etc)
and so you can get both drives fail within a few days of each other.
So: I will buy 2 disks from different manufacturers in the hope that they will
fail at different times.
--
Alain Williams
Linux/GNU Consultant - Mail systems, Web sites, Networking, Programmer, IT Lecturer.
+44 (0) 787 668 0256 https://www.phcomp.co.uk/
Parliament Hill Computers. Registration Information: https://www.phcomp.co.uk/Contact.html
#include <std_disclaimer.h>
Reply to: