Hey All,
What's up? Where are you guys? "Guys" reminded me... are there any women
on this list (I'm waiting for the "you're the only woman I know on the
list" retort)?
I'm having trouble getting my terabyte server off the ground. The Seagate
drives I speced were giving my Sil 3114R SATA controller hell. Under
light load they would work fine, but when I stressed the system (i.e.
building a software RAID 5) they would choke, and throw ATA errors. The
errors were related to buffer timeouts. I found another person having
similar problems with a Sil 3112 controller and Seagate SATA drives, he
found no resolution. So... I sent all eight drives back in exchange for
Maxtor disks. Hopefully the same problem won't occur with the new drives
when they arrive.
I have two SATA chipsets on my motherboard that provide 4 channels each
(Sil 3114R and Nvidia nForce 4 Pro). I've created stable software RAID 5s
on both chipsets outside of my office, but I've been using different
disks. Problems weren't related to a single disk, or a couple of disks.
All disks connected to the Sil 3114 chipset caused ATA timeout errors when
under high load (I tested each individually) -- which really sucked when
building a RAID 5, as about halfway through the build process a disk would
begin timing out, resulting in the RAID compensating for the disk failure,
then, shortly after, another disk would timeout and blow the whole RAID
away. Each disk on the nVidia controller worked fine -- I was able to
sucessfully establish a software RAID across the disks attached to said
controller, and each disk passed a destructive badblock write/read test.
*ugh* headaches! Anyone else experience this, or a related problem?
- Sebastian
_______________________________________________
RLUG mailing list
[email protected]
http://lists.rlug.org/mailman/listinfo/rlug
