![[NetBSD Logo]](/tnf/resource/NetBSD-headerlogo.png)
What to do when you run out of (ancient) 50 pin SCSI disks?
Recently on a few mailing lists a discussion about creating a simple turbo-channel USB adapter came up. I have an old Alpha machine with turbo channel and no on-board USB, so considered ordering one - but wanted to know upfront whether that alpha machine still worked.
I started with updating the installation on the alpha to -current, and ran out of disk space!
No big deal, it still had the original ~1GB disk (RZ26) in it, and a full NetBSD install nowadays needs slightly more. I have plenty of unused SCSI disks lying around, but when looking for 50pin ones, I found I have none left.
Another machine I was looking at for a different root disk solution was an old mac68k (Quadra 640 AV). I had replaced the disk it came with by an 8GB one, but it draws too much power for the internal PSU, so I had to move it to an external SCSI enclosure, which is inconvenient.
Additionally both disks made quite some noise and probably would not last a lot longer. So I started looking for "cheap", silent and low-power alternatives.
I found the SCSI2SD project, which looked promising and had a firmware update earlier this year. I ordered a few boards from idead studio and started testing.
The solution is mostly plug & play, a configuration utility allows updating the firmware and configuring the devices.
I did several bonnie++ runs to get an idea of the speed achievable. All machines are not fast, and disk access has always been slow. To get an overall idea, I also tested on a faster alpha (still with slow scsi disks). Then I moved both test machines over to "root on sd card" and tested again. The results look like this:
| Test | Sequential Output | Sequential Input | Random Seeks | Sequential Create | Random Create | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Per Chr | Block | Rewrite | Per Chr | Block | K/s | CPU | Create | Read | Delete | Create | Read | Delete | ||||||||||||
| K/s | CPU | K/s | CPU | K/s | CPU | K/s | CPU | K/s | CPU | /s | CPU | /s | CPU | /s | CPU | /s | CPU | /s | CPU | /s | CPU | |||
| Ref: DS20 | 3436 | 35 | 3460 | 6 | 2935 | 5 | 9565 | 94 | 19494 | 285.7 | 16 | 3 | 878 | 81 | 31793 | 99 | 2939 | 54 | 940 | 84 | 1351 | 96 | 203 | 10 |
| Alpha HD | 906 | 92 | 2146 | 41 | 980 | 24 | 947 | 93 | 2178 | 23 | 78.5 | 14 | 36 | 95 | 926 | 72 | 290 | 90 | 37 | 96 | 39 | 97 | 25 | 34 |
| Alpha SD | 548 | 77 | 897 | 19 | 450 | 9 | 490 | 97 | 1140 | 9 | 24.9 | 4 | 85 | 89 | 1245 | 75 | 437 | 84 | 90 | 93 | 119 | 96 | 29 | 16 |
| Mac HD | 98 | 95 | 737 | 89 | 373 | 91 | 94 | 96 | 647 | 89 | 14.1 | 47 | 22 | 93 | 502 | 75 | 122 | 81 | 23 | 94 | 31 | 97 | 14 | 43 |
| Mac SD | 95 | 95 | 646 | 89 | 339 | 92 | 93 | 96 | 636 | 79 | 14.5 | 75 | 22 | 90 | 473 | 75 | 124 | 85 | 23 | 92 | 32 | 97 | 13 | 46 |
So even for machines with very slow hard disk access, there is a (relatively small?) performance hit with the SCSI2SD solution. Whether it is too bad for your application, depends on various factors - all of the affected machines are very slow overall anyway.
Just a few days ago a new hardware revision of the SCSI2SD cards has been announced which is supposed to improve performance in the future (when the firmware has caught up), so it may be worth to wait (or get the newer hardware) if performance is important.
I have a VAX machine that could go with this solution as well, but results are expected to be similar (or worse) than for the slow alpha, as my vax is pretty fast (for a vax) and has a decent SCSI subsystem.
For reference, below are the full dmesg from the fast reference and both test machines after switching to SD root, and the sniplet for the original hard disk
AlphaServer DS20
"Fast" reference machine
file system: FFSv2, log
Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016
The NetBSD Foundation, Inc. All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
NetBSD 7.99.29 (GENERIC-$Revision: 1.368 $) #7: Mon May 23 20:44:04 CEST 2016
martin@martins.aprisoft.de:/ssd/src/sys/arch/alpha/compile/GENERIC
AlphaServer DS20 500 MHz, s/n AY94910150
8192 byte page size, 1 processor.
total memory = 1024 MB
(2712 KB reserved for PROM, 1021 MB used by NetBSD)
avail memory = 994 MB
timecounter: Timecounters tick every 0.976 msec
Kernelized RAIDframe activated
mainbus0 (root)
cpu0 at mainbus0: ID 0 (primary), 21264-4
cpu0: Architecture extensions: 0x303
tsc0 at mainbus0: 21272 Core Logic Chipset, Cchip rev 0
tsc0: 8 Dchips, 2 memory buses of 32 bytes
tsc0: arrays present: 512MB, 512MB, 0MB, 0MB, Dchip 0 rev 1
tsp0 at tsc0
pci0 at tsp0 bus 0
pci0: i/o space, memory space enabled, rd/line, rd/mult, wr/inv ok
sio0 at pci0 dev 5 function 0: vendor 1080 product c693 (rev. 0x00)
cypide0 at pci0 dev 5 function 1: Cypress 82C693 IDE Controller (rev. 0x00)
cypide0: bus-master DMA support present, but unused (registers at unsafe address 0x10000)
cypide0: primary channel wired to compatibility mode
cypide0: primary channel interrupting at isa irq 14
atabus0 at cypide0 channel 0
cypide1 at pci0 dev 5 function 2: Cypress 82C693 IDE Controller (rev. 0x00)
cypide1: hardware does not support DMA
cypide1: primary channel wired to compatibility mode
cypide1: secondary channel interrupting at isa irq 15
atabus1 at cypide1 channel 0
ohci0 at pci0 dev 5 function 3: vendor 1080 product c693 (rev. 0x00)
ohci0: interrupting at isa irq 10
ohci0: OHCI version 1.0, legacy support
usb0 at ohci0: USB revision 1.0
isa0 at sio0
lpt0 at isa0 port 0x3bc-0x3bf irq 7
com0 at isa0 port 0x3f8-0x3ff irq 4: ns16550a, working fifo
com0: console
com1 at isa0 port 0x2f8-0x2ff irq 3: ns16550a, working fifo
pckbc0 at isa0 port 0x60-0x64
attimer0 at isa0 port 0x40-0x43
pcppi0 at isa0 port 0x61
midi0 at pcppi0: PC speaker
spkr0 at pcppi0
isabeep0 at pcppi0
fdc0 at isa0 port 0x3f0-0x3f7 irq 6 drq 2
mcclock0 at isa0 port 0x70-0x71: mc146818 compatible time-of-day clock
attimer0: attached to pcppi0
tsp1 at tsc0
pci1 at tsp1 bus 0
pci1: i/o space, memory space enabled, rd/line, rd/mult, wr/inv ok
tlp0 at pci1 dev 7 function 0: Macronix MX98713 Ethernet, pass 0.0
tlp0: interrupting at dec 6600 irq 47
tlp0: Ethernet address 00:40:05:50:ee:9b
tlp0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX
isp0 at pci1 dev 8 function 0: QLogic 1020 Fast Wide SCSI HBA
isp0: interrupting at dec 6600 irq 43
isp1 at pci1 dev 9 function 0: QLogic 1020 Fast Wide SCSI HBA
isp1: interrupting at dec 6600 irq 39
tsciic0 at tsc0
iic0 at tsciic0: I2C bus
timecounter: Timecounter "clockinterrupt" frequency 1024 Hz quality 0
timecounter: Timecounter "PCC" frequency 499882560 Hz quality 1000
scsibus0 at isp0: 16 targets, 8 luns per target
scsibus1 at isp1: 16 targets, 8 luns per target
scsibus0: waiting 2 seconds for devices to settle...
scsibus1: waiting 2 seconds for devices to settle...
uhub0 at usb0: vendor 1080 OHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
sd0 at scsibus0 target 0 lun 0: <COMPAQ, BB00911CA0, 3B05> disk fixed
sd0: 8678 MB, 5273 cyl, 20 head, 168 sec, 512 bytes/sect x 17773524 sectors
sd0: sync (50.00ns offset 8), 16-bit (40.000MB/s) transfers, tagged queueing
sd1 at scsibus0 target 1 lun 0: <COMPAQ, BB00912301, B016> disk fixed
sd1: 8678 MB, 5273 cyl, 20 head, 168 sec, 512 bytes/sect x 17773524 sectors
sd1: sync (50.00ns offset 8), 16-bit (40.000MB/s) transfers, tagged queueing
sd2 at scsibus0 target 2 lun 0: <COMPAQ, BB00911CA0, 3B05> disk fixed
sd2: 8678 MB, 5273 cyl, 20 head, 168 sec, 512 bytes/sect x 17773524 sectors
sd2: sync (50.00ns offset 8), 16-bit (40.000MB/s) transfers, tagged queueing
fd0 at fdc0 drive 0: 1.44MB, 80 cyl, 2 head, 18 sec
sd3 at scsibus0 target 3 lun 0: <COMPAQ, BB00912301, B016> disk fixed
sd3: 8678 MB, 5273 cyl, 20 head, 168 sec, 512 bytes/sect x 17773524 sectors
sd3: sync (50.00ns offset 8), 16-bit (40.000MB/s) transfers, tagged queueing
cd0 at scsibus1 target 5 lun 0: <DEC, RRD47 (C) DEC, 1206> cdrom removable
raid0: RAID Level 1
raid0: Components: /dev/sd0a /dev/sd1a
raid0: Total Sectors: 17773440 (8678 MB)
raid1: RAID Level 1
raid1: Components: /dev/sd2a /dev/sd3a
raid1: Total Sectors: 17773440 (8678 MB)
root on raid0a dumps on raid0b
root file system type: ffs
kern.module.path=/stand/alpha/7.99.29/modules
Alpha DEC 3000
with SCSI2SD
Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016
The NetBSD Foundation, Inc. All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
NetBSD 7.99.29 (GENERIC-$Revision: 1.368 $) #3: Thu May 5 09:17:11 CEST 2016
martin@martins.aprisoft.de:/ssd/src/sys/arch/alpha/compile/GENERIC
DEC 3000 - M300, 150MHz, s/n
8192 byte page size, 1 processor.
total memory = 98304 KB
(2048 KB reserved for PROM, 96256 KB used by NetBSD)
avail memory = 82352 KB
timecounter: Timecounters tick every 0.976 msec
Kernelized RAIDframe activated
mainbus0 (root)
cpu0 at mainbus0: ID 0 (primary), 21064-1
tcasic0 at mainbus0
tc0 at tcasic0: 12.5 MHz clock
sfb0 at tc0 slot 6 offset 0x2000000: 1280x1024, 8bpp
wsdisplay1 at sfb0 kbdmux 1
wsmux1: connecting to wsdisplay1
ioasic0 at tc0 slot 5 offset 0x0: slow mode
le0 at ioasic0 offset 0xc0000: address 08:00:2b:3c:93:27
le0: 32 receive buffers, 8 transmit buffers
zsc0 at ioasic0 offset 0x100000
vsms0 at zsc0 channel 0
wsmouse0 at vsms0 mux 0
zstty0 at zsc0 channel 1 (console i/o)
zsc1 at ioasic0 offset 0x180000
lkkbd0 at zsc1 channel 0
wskbd0 at lkkbd0 mux 1
wskbd0: connecting to wsdisplay1
zsc1: channel 1 not configured
mcclock0 at ioasic0 offset 0x200000: mc146818 compatible time-of-day clock
bba0 at ioasic0 offset 0x240000
audio0 at bba0: full duplex, playback, capture, mmap
tcds0 at tc0 slot 4 offset 0x0: TurboChannel Dual SCSI (baseboard)
asc0 at tcds0 chip 0: NCR53C94, 25MHz, SCSI ID 0
scsibus0 at asc0: 8 targets, 8 luns per target
timecounter: Timecounter "clockinterrupt" frequency 1024 Hz quality 0
timecounter: Timecounter "PCC" frequency 150006528 Hz quality 1000
lkkbd0: no keyboard
scsibus0: waiting 2 seconds for devices to settle...
sd0 at scsibus0 target 1 lun 0: <codesrc, SCSI2SD, 1.0> disk fixed
sd0: 3712 MB, 473 cyl, 255 head, 63 sec, 512 bytes/sect x 7603200 sectors
sd0: async, 8-bit transfers
root on sd0a dumps on sd0b
root file system type: ffs
kern.module.path=/stand/alpha/7.99.29/modules
with original RZ26
sd0 at scsibus0 target 3 lun 0: disk fixed
sd0: 1001 MB, 2570 cyl, 14 head, 57 sec, 512 bytes/sect x 2050860 sectors
sd0: sync (200.00ns offset 15), 8-bit (5.000MB/s) transfers, tagged queueing
Mac Quadra 640AV
File system: FFSv1, log
with SCSI2SD
Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016
The NetBSD Foundation, Inc. All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
NetBSD 7.99.29 (MAC-BETH) #18: Tue May 10 05:40:16 CEST 2016
martin@night-owl.duskware.de:/usr/src/sys/arch/mac68k/compile/MAC-BETH
Apple Macintosh Centris 660AV (68040)
cpu: delay factor 800
fpu: mc68040
total memory = 45056 KB
avail memory = 40248 KB
timecounter: Timecounters tick every 16.666 msec
mrg: 'Quadra AV ROMs' ROM glue, tracing off, debug off, silent traps
mainbus0 (root)
obio0 at mainbus0
esp0 at obio0 addr 0: address 0x2da000: NCR53C96, 25MHz, SCSI ID 7
scsibus0 at esp0: 8 targets, 8 luns per target
adb0 at obio0
intvid0 at obio0 @ 50100600: CIVIC video subsystem
intvid0: 640 x 480, 65536 color
macfb0 at intvid0
wsdisplay0 at macfb0 (kbdmux ignored)
mc0 at obio0: address 08:00:07:46:74:0a
zsc0 at obio0 chip type 0
zsc0 channel 0: d_speed 9600 DCD clk 0 CTS clk 0
zstty0 at zsc0 channel 0 (console i/o)
zsc0 channel 1: d_speed 9600 DCD clk 0 CTS clk 0
zstty1 at zsc0 channel 1
nubus0 at mainbus0
timecounter: Timecounter "clockinterrupt" frequency 60 Hz quality 0
timecounter: Timecounter "VIA1 T2" frequency 783360 Hz quality 100
scsibus0: waiting 2 seconds for devices to settle...
adb0 (direct, Cuda): 2 targets
aed0 at adb0 addr 0: ADB Event device
akbd0 at adb0 addr 2: keyboard II (ISO layout)
wskbd0 at akbd0 (mux ignored)
ams0 at adb0 addr 3: 1-button, 100 dpi mouse
wsmouse0 at ams0 (mux ignored)
sd0 at scsibus0 target 0 lun 0: <codesrc, SCSI2SD, 4.6> disk fixed
sd0: 15104 MB, 1925 cyl, 255 head, 63 sec, 512 bytes/sect x 30934016 sectors
sd0: async, 8-bit transfers
cd0 at scsibus0 target 3 lun 0: <SONY, CD-ROM CDU-8003A, 1.9a> cdrom removable
cd0: async, 8-bit transfers
boot device: sd0
root on sd0a dumps on sd0b
root file system type: ffs
kern.module.path=/stand/mac68k/7.99.29/modules
with hard disk
sd0 at scsibus0 target 0 lun 0: disk fixed
sd0: 8761 MB, 14384 cyl, 3 head, 415 sec, 512 bytes/sect x 17942584 sectors
sd0: async, 8-bit transfers, tagged queueing
[4 comments]
Posted by butts on May 27, 2016 at 06:58 PM UTC #
Posted by tobutts on May 28, 2016 at 11:05 PM UTC #
Posted by Martin Husemann on May 31, 2016 at 10:23 AM UTC #
Posted by Chris on June 15, 2016 at 01:42 PM UTC #