We're running 4.6.2. We had no remote access. The failure was on the swap
partition and it would not respond to TCP connections. The physical console
was still fine.
The worst of this is we tested hot swapping drives, rebuilding arrays in the
background, everything and were thinking we were in good shape. We have a
half-dozen of these boxes with identical hardware configuration in
production now and cannot afford to have them lock up if a drive fails.
I agree about IDE-RAID, but we are on a tight budget and need only
reliability, not performance, from these machines.
On 12/9/02 2:31 PM, "Alexander Langer" <email@example.com> wrote:
> Thus spake Michael DeMan (firstname.lastname@example.org):
>> We are using Promise RAID controllers built on motherboards. These provide
>> simple RAID-1 mirroring for us.
> Same here.
>> My assumption would be that the RAID controller would respond and
>> automatically detach a bad disk.
> That's actually what it should do, and once already did here -- on a
> 4.6-STABLE box - older kernel's showed a similar behaviour. I remotely
> had the impression the box hung, but maybe it had the same symptoms as
> your box. What version are you using?
> Anyways, I wouldn't actually recommend User-Hardware such as the
> ATA-RAID-Controllers for production systems anyways.
Michael F. DeMan
Director of Technology
OpenAccess Internet Services
1305 11th St., 3rd Floor
Bellingham, WA 98225
Tel 360-647-0785 x204
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hardware" in the body of the message