FastSCSI on WarpEngine
I've some disks hooked up to my A3000:
scsi-ncr53c7xx : NCR53c710 at memory 0x40040000, io 0x0, irq 12
scsi0: Revision 0x1
scsi0 : NCR code relocated to 0xbefe5e0 (virt 0x03efe5e0)
scsi0 : test 1 started
wd33c93-1: chip=WD33c93A/9 no_sync=0xff no_dma=0 debug_flags=0x00
setup_args=,,,,,,,,,
Version 1.25 - 09/Jul/1997, Compiled Jul 6 2002 at 21:57:12
scsi0 : Amiga NCR53c710 SCSI
scsi1 : Amiga 3000 built-in SCSI
scsi : 2 hosts.
scsi0 : target 1 accepting period 200ns offset 8 5.00MHz synchronous SCSI
scsi0 : setting target 1 to period 200ns offset 8 5.00MHz synchronous SCSI
Vendor: HP Model: C3725S Rev: 6039
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sda at scsi0, channel 0, id 1, lun 0
scsi0 : target 2 accepting period 200ns offset 8 5.00MHz synchronous SCSI
scsi0 : setting target 2 to period 200ns offset 8 5.00MHz synchronous SCSI
Vendor: HP Model: C3725S Rev: 6039
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sdb at scsi0, channel 0, id 2, lun 0
sending SDTR 0103015e00sync_xfer=30 Vendor: SEAGATE Model: ST32171N
Rev: 0280
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sdc at scsi1, channel 0, id 0, lun 0
sending SDTR 0103013200sync_xfer=20 Vendor: SEAGATE Model: ST34573N
Rev: 5958
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sdd at scsi1, channel 0, id 2, lun 0
sending SDTR 0103015e00sync_xfer=30 Vendor: SGI Model: SEAGATE
ST31230N Rev: 0272
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sde at scsi1, channel 0, id 3, lun 0
sending SDTR 0103015e00sync_xfer=30 Vendor: SEAGATE Model: ST43400N
Rev: 1028
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sdf at scsi1, channel 0, id 4, lun 0
sending SDTR 0103015e00sync_xfer=30 Vendor: IBM Model: DPES-31080
Rev: S31Q
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sdg at scsi1, channel 0, id 5, lun 0
scsi : detected 7 SCSI disks total.
SCSI device sda: hdwr sector= 512 bytes. Sectors= 4194058 [2047 MB] [2.0 GB]
SCSI device sdb: hdwr sector= 512 bytes. Sectors= 4194058 [2047 MB] [2.0 GB]
SCSI device sdc: hdwr sector= 512 bytes. Sectors= 4222640 [2061 MB] [2.1 GB]
SCSI device sdd: hdwr sector= 512 bytes. Sectors= 8888924 [4340 MB] [4.3 GB]
SCSI device sde: hdwr sector= 512 bytes. Sectors= 2070235 [1010 MB] [1.0 GB]
SCSI device sdf: hdwr sector= 512 bytes. Sectors= 5688447 [2777 MB] [2.8 GB]
SCSI device sdg: hdwr sector= 512 bytes. Sectors= 2118144 [1034 MB] [1.0 GB]
As you can see above sda and sdb are connected to the WarpEngine, but only
with 5 MHz transfers.
Ok, regarding to /usr/src/linux/drivers/scsi/53c7xx.c:
* If you really want to run 10MHz FAST SCSI-II transfers, you should
* know that the NCR driver currently ignores parity information. Most
* systems do 5MHz SCSI fine. I've seen a lot that have problems faster
* than 8MHz. To play it safe, we only request 5MHz transfers.
*
* If you'd rather get 10MHz transfers, edit sdtr_message and change
* the fourth byte from 50 to 25.
(http://www2.randomlogic.com/linux_html/400601.html)
Usually I only get around 2 MB/s from disks, either via scsi0 or scsi1.
Using a raid0 on scsi0 *or* scsi0 doesn't improve speed at all, but raid0 on
scsi0 *and* scsi1 do (of course, using sdb and sdc), giving a total speed of
roughly 3 MB/s, which is not as speedy as it could be, even with 5 MHz
transfers instead of 10 MHz.
So, the question is:
Has someone tried to run the NCR on the WarpEngine with 10 MHz transfers
successfully? All disks are 7200 rpm except the Seagate ST43400N (sdf), so I
guess the disks are fast enough.
Regarding to the above text from kernel source, it might get parity errors
with 10 MHz, but how "dangerous" is this? Disks sda, sdb (scsi0), and sdc
(scsi1) are hooked up internally in the A3000 (yes, desktop ;)), and the
cable length is below 50 cm. All other disks are within an external 4x SCSI
tower.
How possible are parity errors? Will it result in data corruption? Wouldn't
be that good when arrakis would produce corrupted Debian packages, I
guess... ;))
--
Ciao... //
Ingo \X/
Reply to: