[jggimi]

I don't often ask for OpenBSD administrative advice, I usually provide it...whether or not it is correct advice.

One of my servers uses RAIDframe in a root-on-RAID configuration. I've had bad sectors on one of the drives multiple times, and with RAIDframe have had no unscheduled downtime. I've taken drives out of service, "repaired" bad sectors by having the drive replace them from spare, and placed them back into service, without rebooting. No sector failures in the last 18 months, though.

I use Softraid on my netbook, with the CRYPTO discipline, to encrypt /home at rest. I use it with a DUID so that it mounts regardless of sd # assignment. I like Softraid, it is easier to provision and manage than RAIDframe. But it is not quite yet "production RAID" ready, as root-on-RAID is not yet possible. Altroot provides a method to recover from a failed root drive, but the root drive remains a single point of failure.

That said, root-on-RAID is coming to -current. Soon. I am thinking about becoming an early adopter. As such, I want to know if I'll be in the minority, or among the majority.

An informal poll is attached to this thread, with questions for the community regarding your administration of software RAID implementations. If you administrate any OpenBSD system, even if its just a single workstation for personal use, please respond to the poll, whether or not you use any RAID implementations at all.

I'm also seeking advice from anyone who has gone down the path of softraid for RAID management (as opposed to CRYPTO) already, before root-on-RAID is available.

Thanks!

Edit: You may select multiple choices, if you manage multiple RAID environments.
Edit: please consider storage that is an externally managed RAID subsystem equivalent to BIOS-managed, when responding to the poll.