On Tue, 30 May 2000, Juri Haberland wrote:
> Luca Berra wrote:
> >
> > get the latest lilo from metalab, it understands raid
>
> But it does not really work.
> If I do a lilo with boot=/dev/md0 in my lilo.conf it does write to both
> disks. But booting from the second disk in the array is not p
On Sat, 20 May 2000, Mika Kuoppala wrote:
> For all users of my read1balance patch, please
> repatch against Ingo's 2.2.15-A0 patch
Does this mean raid1readbalance-2.2.15-1 has changed again (after May 16,
when we changed MAX_SINGLE_DISK_READS to 500)?
> There was a change that your whole mirr
On Fri, 19 May 2000, Martin Bene wrote:
> Since you're into comapring benchmarks, let me contribute something from an
> IBM Netfinity server with IBM ServeRaid 3L Controller, Firmware + Kernel
> driver 3.60. Server is running with 1 Coppermine 650 Cpu, Kernel is 2.2.15
RAID5?
> Dir Size Bl
On Fri, 19 May 2000, Ard van Breemen wrote:
> On Wed, May 17, 2000 at 07:19:43PM -0300, Christian Robottom Reis wrote:
> > If you want to know what slow means, I'll post some SW-Raid readbalanced
> > RAID1 benchmarks *grin* Mika rules!
> (*) Yes! I want to know what slow
On Wed, 17 May 2000, Thomas King wrote:
> Do anybody have some experiencies with ICP VORTEX or MYLEX Hardware-RAID
> controller?
> Which controller vendor should I prefer?
I had a great time with an Acceleraid 250 a while back. Performance was
quite sad for an LVD Ultra2 SCA Raid1 with two drive
On Sat, 8 Apr 2000, Ed Carp wrote:
> There are very few problems with software that can cause hardware to
> die. The *only* time I've heard of this happening is if you get the
> sync rate wrong on your monitor. It's an axiom of programming that is
> occasionally not understood by non-geeks that
On Fri, 7 Apr 2000, Ed Carp wrote:
> > Does this smell the most of hardware or software? Can the nfsv3 patches be
> > so evil? Is driver 2.2.5 silently corrupting board firmware?
>
> Sounds like hardware problems. What happened when you tried to
> reflash the firmware?
Well, I erased the confi
On Fri, 7 Apr 2000, Dan Jones wrote:
> I examined a number of recovery cases and what you propose to do
> is well supported by Mylex. First, the controller defers to the
> configuration information on the drives when there is a conflict
> with the information stored in the controller. Makes sens
Just a short note to aid me in my recovery here: can I remove the disks
from a working array, reset the board configuration, put them back again,
and rebuild the configuration - with no problem? Should I also backup the
configuration to a disk?
I'm wondering if I can swap drives from the dead bo
I'm running the Mylex driver 2.2.5 on a two-disk RAID1 production box
(racked here on an ISP). I've changed the firmware to 4.07-07 listed on
the page for the driver, and stress-tested it for a couple of days before
shipping.
The setup was a bit flaky to start off with, but the firmware switch a
On Wed, 15 Mar 2000, Luigi Gangitano wrote:
> I need some help setting up boot on a RAID1 device. I used Method 2 of
> Jakob OEstergaard's latest HowTo. I got it working, now my system mounts
> /dev/md0 as root partition. But if I try to update LILO (to make it working on
> the second hd's MB
On Wed, 15 Mar 2000 [EMAIL PROTECTED] wrote:
> > Can you please tell me where can I find the latest raid-tools? I
> found raidtools- > dangerous on http://people.redhat.com/mingo and
> didn't try to use it.
>
> Good question. From my impression, Linux (software) raid support got
> a bit divided
On Tue, 14 Mar 2000, Malcolm Beattie wrote:
> Benchmarking it on a stripeset of 7 x 9GB disks on a Ultra3 bus with
> one of the Adaptec 7899 channels, it's impressively fast. 81MB/s block
> reads and 512 seeks/s in bonnie and 50MB/s (500 "netbench" Mbits/sec)
> running dbench with 128 threads. I'
Just FYI, a run on a Netfinity 5000 with a ServeRAID card and two IBM 8G
LVD disks plugged into a backplane. I can dig up the model if it makes
things more meaningful. mem=16M, runlevel 1, numruns 5.. you know the
drill. AFAICS to me the ServeRAID is LVD as well, which should give us
80Mb/s max t
On Thu, 9 Mar 2000, Holger Kiehl wrote:
> > a) Make RAID bootdisk.
> > b) boot up and mkraid
> >
> To do the mkraid you need a raidtab file and for that you need an
> editor or it must be copied to the floppy when you create it.
Certainly, but Slackware is nice in that it lets you do whatever yo
On Tue, 7 Mar 2000, Peter Palfrader wrote:
> Would be nice if you could send me those files.
> (If it's larger than 2meg please send it to [EMAIL PROTECTED])
Sure. I trust you can parse the xspread data out if you want it - I just
wanted to calculate variances so I used it right there. You can n
Sorry. Not ~mingo
http://people.redhat.com/mingo/raid-patches/
Cheers,
--
_/\ Christian Reis is sometimes [EMAIL PROTECTED]
\/~ suicide architect | free software advocate | mountain biker
Did the superblock formats change, or is it okay to boot a kernel with
new-style support on an old-style array? Simmetrically, can arrays
created with raidtools-0.90 be booted on old-style kernels?
I'm wondering on the forced upgrade path on my Slackware patch. I'm not
really worried about break
On Wed, 8 Mar 2000, Saibot wrote:
> I'm rather new to the linux world (only a year since I first
> put my hands in this) and I'm now assigned the task to maintain a server.
> I'm right now having a problem with RAID (software raid that is). it
> didn't work with the previous versions so I tried w
We're working on a patch that might make it's way into the next Slack
release. In the meantime, I can suggest you do it completely differently:
a) Make RAID bootdisk.
b) boot up and mkraid
c) modify 'setup' so it understands your md drives
d) install away as if nothing was different
e) boot with
On Thu, 9 Mar 2000, Jakob Østergaard wrote:
> On Wed, 08 Mar 2000, Brian Pomerantz wrote:
>
> > On Thu, Mar 09, 2000 at 12:44:32AM +0100, Jakob Østergaard wrote:
> > >
> > If there isn't hot-swap RAID 5 with auto rebuild, it will never
> > happen.
>
> It would be nice if a program such as ASCI
On Thu, 9 Mar 2000, Frank Joerdens wrote:
> I also tried patching a 2.0.36, a 2.2.14 and a 2.2.12 kernel, all with
> similar results.
correct patches and tools @ people.redhat.com/mingo/raid-patches
Cheers,
--
_/\ Christian Reis is sometimes [EMAIL PROTECTED]
\/~ suicide architect | free softw
I had a look at the numbers you got on the benchmark you posted, and I've
tried averaging out the values to see if -Rstripe made any difference.
FWIW, it seems there's a small (but existant, IMHO) improvement on reads
and not much change on writes. Reads seemed more improved running 2+
threads, p
Same setup, but with kernel 2.2.14 with the vanilla RAID patch from
~mingo. Not much difference at all from the former - notice I've run it
with a couple of extra threads (16,32,64), and more threads get more seeks
(but worse throughput).
Chunk is 4k, Stride is 4 and Block is 1024k
Machine
I've run a couple of benchmarks using different numruns on the same
hardware, calculated the variations and deviations, and found that numruns
should be 4 or more on my current setup [P3, linux-2.2.14-Raid1, mem=16M,
runlevel 1, single AHA2940UW1, two Quantum Atlas IV] to keep variance down
enoug
[I tried to apply the low-latency patch as well, but it panicked all over
the place and I gave up.]
Same setup, notice the throughput and seek improvement over the last one
with numthreads > 1. The difference the block size makes over throughput
and seek grows a bit. Chunk doesn't do much. I lik
> # tiobench.pl --numruns 5 --size 1024
> Size is MB, BlkSz is Bytes, Read and Write are MB/sec, Seeks are Seeks/sec
>
> Dir Size BlkSz Thr# Read (CPU%) Write (CPU%) Seeks (CPU%)
> - -- --- - -- --
> .1024 40961 25.6001
Running on a P3-550, with mem=48M, Redhat 6.1 stock raid kernel-2.2.12-20,
with an AHA2940UW and two Quantum Atlas IV running RAID1. Runlevel 1 and
nothing running apart from tiobench.
Notice the improvement in throughput and seek with growing block-sizes [a
device of the benchmark, perhaps?] an
On Thu, 2 Mar 2000 [EMAIL PROTECTED] wrote:
> I think the bonnie test at least tells me what max throughput of the
> drives and controller's ability to do RAID-5 are. I'll be happy to run
> other benchmarks though. Where can I find tiotest? Searches on
> google/altavista/freshmeat turned up no
On Thu, 2 Mar 2000 [EMAIL PROTECTED] wrote:
> On Wed, 1 Mar 2000, Leon Brouwers wrote:
>
> > * DAC960 RAID Driver Version 2.2.4 of 23 August 1999 *
> > Copyright 1998-1999 by Leonard N. Zubkoff <[EMAIL PROTECTED]>
> > Configuring Mylex DAC1164P PCI RAID Controller
> > 0:1 Vendor: WD
James, when run tiotest with a size too small for the number of threads
I'm testing, I get a rather cryptic
Error in seek/read, off=0, read=0, seeks=0 : : Success
Error in seek/read, off=0, read=0, seeks=0 : : Success
which tiotest.c emits at line 616.
It happens when I supply a file size that
On Tue, 29 Feb 2000, James Manning wrote:
> Tell ya what, pick out an isolated case which is heavily reproducible,
> print out the tiobench output, then print out the tiotest output.
Never mind. It seems I hadn't installed the new kernel [I swear I checked]
after all, and thereforce the numbers
On Tue, 29 Feb 2000, James Manning wrote:
> [ Tuesday, February 29, 2000 ] Christian Robottom Reis wrote:
> > I've got the simple scripts I used to do the benchmarks here and if
> > somebody wants to have a look, feel free.
>
> go ahead and mail them to the list as at
On Wed, 1 Mar 2000, James Manning wrote:
> per-char doesn't matter (one of the reasons I hate ppl using bonnie,
> besides the single-threaded-ness). Considering the queueing and scat/gat
Why not? Because usual disk operations are done block by block?
Cheers,
--
_/\ Christian Reis is sometimes [
I've seen a lot of variation on various runs of tiotest using the same
setup - even in single-user mode. Is this expected, and do you know why it
happens? Is it just the effect of the buffer cache, or do we avoid using
it?
What's a decent --numruns to use, taking into evidence such
variation? I'
On Tue, 29 Feb 2000, Leonard N. Zubkoff wrote:
> > * DAC960 RAID Driver Version 2.2.4 of 23 August 1999 *
> > Copyright 1998-1999 by Leonard N. Zubkoff <[EMAIL PROTECTED]>
> > Configuring Mylex DAC960PG PCI RAID Controller
> > Firmware Version: 4.06-0-08, Channels: 1, Memory Siz
On Tue, 29 Feb 2000, James Manning wrote:
> [ Tuesday, February 29, 2000 ] Christian Robottom Reis wrote:
> > /proc/rd/ relevant information:
> >
> > * DAC960 RAID Driver Version 2.2.4 of 23 August 1999 *
> > Copyright 1998-1999 by Leonard N. Zubkoff <[EMA
On Mon, 14 Feb 2000, Peter Pregler wrote:
> All is fine but during reconstruction I get a few syslog-messages that I
> simply cannot believe are true. The message in question are:
>
> Feb 12 11:31:52 kludge kernel: md: serializing resync, md8 has overlapping
> physical units with md9!
Just mean
On Sun, 13 Feb 2000, Johan Ekenberg wrote:
> 0.90. Every server has a Raid-5 array consisting of 5 large IBM scsi disks +
> one spare. It works like a charm, extremely fast and no trouble at all with
How fast are the IBM disks? We're using Quantums here and they suck!
> Software-RAID during mon
James, I've run a whole truckload of benchmarks on raid1 with varying
chunksizes on three different kernels, and on a plain disk. I'm about to
publish some of the stuff, but I'm wondering very hard why is it that the
readbalancing test showed _awful_ numbers on tiotest 0.21 and great
numbers on 0
/proc/rd/ relevant information:
* DAC960 RAID Driver Version 2.2.4 of 23 August 1999 *
Copyright 1998-1999 by Leonard N. Zubkoff <[EMAIL PROTECTED]>
Configuring Mylex DAC960PG PCI RAID Controller
Firmware Version: 4.06-0-08, Channels: 1, Memory Size: 4MB
PCI Bus: 0, Device: 10, Funct
I've run about 100 hours of benchmarks for the last days, and I'll say the
following: chunksize is probably meaningless on raid1 if my data is
anything close to consistent, and so (or therefore?) is supplying -Rstripe
to mke2fs. The block-size, however, is 100% significant. I've got the
simple sc
42 matches
Mail list logo
