Here's a link to the undeadly.org story.
Original email from Theo to misc@.
Basically, OpenBSD has some neat raid management. It can utilize a hot spare and rebuild degraded arrays without going down. bioctl(8) can help you find the disk that's bad, mess with the alarm, and designate hot spares.
Support for some enclosure monitors is working too, and can be queried or monitored by sensorsd(8).
Support is only available for ami(4) controllers now though. Hopefully we'll see more support in the future. Maybe even Adaptec will reconsider their silliness.
Original email from Theo to misc@.
Basically, OpenBSD has some neat raid management. It can utilize a hot spare and rebuild degraded arrays without going down. bioctl(8) can help you find the disk that's bad, mess with the alarm, and designate hot spares.
Support for some enclosure monitors is working too, and can be queried or monitored by sensorsd(8).
Support is only available for ami(4) controllers now though. Hopefully we'll see more support in the future. Maybe even Adaptec will reconsider their silliness.
