expensive and relies on closed software. SCSI solutions can be neater
(e.g., use 3 of the Enlight EN-8700 5-in-3 cases for 15 drives in the same
space) and have better performance and possibly robustness, but also cost more.
Jan Edler
NEC Research Institute
than a problem with the
raid layer per se.
Jan Edler
NEC Research Institute
Benno Senoner wrote:
I was wondering how much IDE channels linux 2.2 can handle,
can it handle 8 channels ?
I think the limit with the later 2.2 kernel ide patches is 10 IDE channels.
I have run quite a bit with 4 Promise cards (8 channels),
plus the 2 onboard PIIX channels.
Jan Edler
NEC
On Tue, Jan 11, 2000 at 04:25:27PM +0100, Benno Senoner wrote:
Jan Edler wrote:
I wasn't advising against IDE, only against the use of slaves.
With UDMA-33 or -66, masters work quite well,
if you can deal with the other constraints that I mentioned
(cable length, PCI slots, etc).
Do
that support udma-66 (or at least
udma-33). This allows you to recover rather more quickly
from a drive failure, assuming you buy at least 1 extra
hot-swap box and drive. Even if you don't mind rebooting
to deal with a failure, it sure beats tearing open the machine.
Good luck,
Jan Edler
On Mon, Jan 10, 2000 at 12:49:29PM -0800, Dan Hollis wrote:
On Mon, 10 Jan 2000, Jan Edler wrote:
- Performance is really horrible if you use IDE slaves.
Even though you say you aren't performance-sensitive, I'd
recommend against it if possible.
My tests indicate UDMA performs
and they are taking a fairly closed
attitude towards the software. The driver is distributed in binary
form. You apply a bunch of kernel patches, and link in their driver.
This causes all sorts of problems.
Jan Edler
NEC Research Institute
Did something change? The raidzone Smartcans we have
are just nice packages for ATA drives so you can do software raid.
One ATA controller per drive, with nice slim hot-swap carriers.
Jan Edler
NEC Research Institute
.
the big thing is increasing density. The performance boost
is almost secondary, but follows as a direct result of increased density.
Jan Edler
NEC Research Institute
On Mon, Aug 30, 1999 at 09:48:18AM -0400, Mike Frisch wrote:
For example, IBM's Deskstar 25GP and Deskstar 22GXP list
web search.
Jan Edler
NEC Research Institute
on the outer cylinder. Lower numbered blocks are always faster
than higher numbered blocks, in my experience. Of course, there's no
physical reason why it should be so; manufacturers could just as easily
do it the other way. But I appreciate that there appears to be
a convention.
Jan Edler
NEC
ylinders,
and almost 18MB/s on the inner ones. This is the fastest sustained rate
I've personally seen on any drive, ATA or SCSI. I suppose there are
faster drives out there, I just haven't seen them.
But I didn't check them on udma-33.
Jan Edler
NEC Research Institute
saturation is not a problem for 8 drives on 1 bus,
and that something else is limiting performance.
Do you have comparable results for 8 drives on one bus?
Jan Edler
NEC Research Institute
?
Jan Edler
NEC Research Institute
% of one Xeon.
Bonnie just takes cpu/elapsed*100.
Still, it can't be right.
Jan Edler
NEC Research Institute
On Thu, Aug 19, 1999 at 10:39:01PM +0100, Paul Jakma wrote:
On Thu, 19 Aug 1999, Jan Edler wrote:
It isn't reporting 99% of 4 Xeons, but 99% of one Xeon.
Bonnie just takes cpu/elapsed*100.
it also depends on the load. if bonnie is the only real load on the
machine then cpu util
ell suited to tons of random I/Os.
Still, I consider the lack of higher rates to be an unsolved mystery.
Jan Edler
NEC Research Institute
over 2 channels sure indicate to do so :)
Or, as I say, it might indicate the mylex or driver is maxed out.
Jan Edler
--- Bonnie.c.unhacked Wed Aug 28 12:23:49 1996
+++ Bonnie.cFri Aug 6 13:41:27 1999
@@ -148,6 +148,7 @@
size = Chunk * (size / Chunk);
fprintf(stderr, "Fil
and NR_STRIPES
settings later on. I'm getting worried I'm not bottlenecking on anything
scsi-related at all, and it's something else in the kernel *shrug*
I agree, something is wrong to produce 99% utilization in such a situation.
Jan Edler
NEC Research Institute
It's Byte, so 66 Megabytes/sec is correct.
But that's just the burst rate. If you look at the specs
for the disk drives, you'll see that they can't keep up with that rate.
I figure anything 10MB/s sustained is pretty good for todays drives.
Expect more in the future.
Jan Edler
NEC Research
-i /dev/hdX will tell what UDMA mode is enabled.
I suppose you might need to get a utility (probably for dos or windows)
from the drive manufacturer to set the UDMA mode.
Jan Edler
NEC Research Institute
I don't buy this; the atime updates should be subject to caching,
and not get written to the disk more than the update daemon
(kflushd or whatever) forces.
Jan Edler
NEC Research Institute
On Thu, Jul 29, 1999 at 09:20:15AM -0500, Tim Walberg wrote:
For pure reads, there should
know when I'll get
time to try fixing that problem, but for now uniprocessor is ok for
my testing.
On a Pentium II/400 I get ~60MB/s reading a file with raid0 on 6 drives,
but 40MB/s with raid5 on 8 drives.
Thanks,
Jan Edler
NEC Research Institute
23 matches
Mail list logo
