Tuning Linux for high-speed disk subsystems

Tuning Linux for high-speed disk subsystems

Post by Ragnar Kj?rsta » Thu, 15 Nov 2001 01:50:11




> After some testing at Compaq's lab in Oslo, I've come to the conclusion
> that Linux cannot scale higher than about 30-40MB/sec in or out of a
> hardware or software RAID-0 set with several stripe/chunk sizes tried out.

Eh, we do 60-70 MB/s reads and 110-120 MB/s writes on our RAIDs... from
linux.

Quote:> Does anyone know this stuff good enough to help me how to tune the system?
> PS: The CPUs were almost idle during the test. Tested file system was
> ext2.

I'd say you should get rid of your compaq raid controller and use a
regular SCSI-controller - 66Mhz 64 bit. (e.g. an adaptec)

--
Ragnar Kj?rstad
Big Storage
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

 
 
 

Tuning Linux for high-speed disk subsystems

Post by Alan Co » Thu, 15 Nov 2001 01:50:10


Quote:> After some testing at Compaq's lab in Oslo, I've come to the conclusion
> that Linux cannot scale higher than about 30-40MB/sec in or out of a
> hardware or software RAID-0 set with several stripe/chunk sizes tried out.
> The set is based on 5 18GB 10k disks running SCSI-3 (160MBps) alone on a
> 32bit/33MHz PCI bus.

I'm beating that with IDE 8)

Quote:> After speking to the storage guys here, I was told the problem generally
> was that the OS should send the data requests at 256kB block sizes, as the
> drives (10k) could handle 100 I/O operations per second, and thereby could

Right now we tend to queue 128 blocks per write. That can be tuned if you
want to play with it.

Quote:> Does anyone know this stuff good enough to help me how to tune the system?
> PS: The CPUs were almost idle during the test. Tested file system was
> ext2.

Im not sure the best way to get big linear blocks in the ext2 layout or
if perhaps XFS would do that job better, but the physical layer comes
down the the block limit, scsi max sectors per I/O set by the controller
and to an extent the vm readahead (tunable in -ac kernels - the patch
to md.c should tell you how to hack md for that)

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

 
 
 

1. cdrecord: can't write to high-speed CD-RW disk

The situation is this:

Dell OptiPlex, running RH7.2 with 2.4.9 kernel, Creative CD-RW writer.
No trouble writing to Kodak 1x-4x speed CD-RW.
However, when trying to write to Maxell 4x-10x speed CD-RW, neither
blanking nor writing works. However, on a Windows box the same disk
can be blanked and written.

Here's the output of cdrecord:

# cdrecord -v speed=4 blank=all dev=0,0,0
Cdrecord 1.10 (i686-pc-linux-gnu) Copyright (C) 1995-2001 J?rg
Schilling
TOC Type: 1 = CD-ROM
scsidev: '0,0,0'
scsibus: 0 target: 0 lun: 0
Linux sg driver version: 3.1.20
Using libscg version 'schily-0.5'
atapi: 1
Device type    : Removable CD-ROM
Version        : 0
Response Format: 1
Vendor_info    : 'CREATIVE'
Identifikation : ' CD-RW RW8438E  '
Revision       : 'FC03'
Device seems to be: Generic mmc CD-RW.
Using generic SCSI-3/mmc CD-R driver (mmc_cdr).
Driver flags   : SWABAUDIO
Drive buf size : 1359872 = 1328 KB
Current Secsize: 2048
  ATIP start of lead out: 359849 (79:59/74)
Blocks total: 359849 Blocks current: 347849 Blocks remaining: 347999
Starting to write CD/DVD at speed 2 in write mode for single session.
Last chance to quit, starting real write in 0 seconds. Operation
starts.
Blanking entire disk
cdrecord: Input/output error. blank unit: scsi sendcmd: no error
CDB:  A1 00 00 00 00 00 00 00 00 00 00 00
status: 0x2 (CHECK CONDITION)
Sense Bytes: 70 00 05 00 00 00 00 0A 00 00 00 00 27 00 00 00
Sense Key: 0x5 Illegal Request, Segment 0
Sense Code: 0x27 Qual 0x00 (write protected) Fru 0x0
Sense flags: Blk 0 (not valid)
cmd finished after 0.001s timeout 9600s
cdrecord: Cannot blank disk, aborting.

You will notice that the actual write speed is 2x, instead of the 4x
requested on the command line. You will also notice that there's a
"write protected" message in the output. Similar things happen when
trying to write an image to the disk. Why the disk should be
write-protected, I have no idea.

So what's up with this disk, and what can I do to make it usable?

2. Shrink-Wrapped Linux !

3. High Performance Disk Subsystem

4. syslog via email

5. High-speed serial and Linux?

6. How to grow a memory-mapped disk file automatically?

7. Solstice X.25

8. LOCAL: High-speed networking research using Linux

9. AOL high-speed cable access and linux

10. High-Speed Connection on FreeBSD 4.4

11. High-speed modem at built-in serial port (RS/6000)