rlcarr@athena.mit.edu (Rich Carreiro) (11/14/89)
I am looking into buying a hard drive for my A2000. I have heard all sorts of conflicting stuff. So, given some conditions, can anyone give me some recommendations? Just as welcome are brands to stay away from. Ok - what I want: 1) 80+ meg hard drive 2) controller must autoboot 3) preferably can control more than one drive 4) preferably allows drive to be mounted right to card (though not a necessity) 5) preferably an internal drive [also, my A2000 has 1 meg chip RAM, 2 meg (hopefully soon to be 4 meg) fast RAM, and eventually some sort of accelerator board, in the event that any of these things have an impact on what kind of drive to get] Some other questions - I have heard various pro/con arguments about DMA controllers. So, what's the opinion out there? Will 1.4 and/or the ECS change those opinions. My dealer recommends the GVP drive controller. Agree? Disagree? Should I get a SCSI? An ST-506? What is the difference? Assuming I get enough responses, I will email summaries to those who wish them, or post to the net if the interest warrants. Any comments, suggestions, etc., will be GREATLY appreciated! Thanks in advance (ADVthanksANCE :-)! -- Rich Carreiro - Most Biased Boston Celtics Fan! Get well soon, Johnny Most! ARPA: rlcarr@space.mit.edu Welcome back, Larry! UUCP: ...!mit-eddie!space.mit.edu!rlcarr McHale, 6th man supreme! BITNET: rlcarr@space.mit.edu "Follow through!"
rlcarr@athena.mit.edu (Rich Carreiro) (11/22/89)
I have narrowed my selection to either a HardFrame or an A2091 for a
controller, and will probably get a SyQuest rmovable hard drive.
Some questions:
1) Which is faster, HardFrame or A2091?
2) Does the A2091 fix the DMA/overscan slowdown problem of the A2090(a)?
3) I've been told that the SyQuest takes 10-15 seconds to spin up.
Will this cause timeout problems with the HardFrame? The A2091??
[note: I called MicroBotics about this - the person who answered was
clueless as to the answer]
4) Does the HardFrame work with the 6.2 motherboard?
^
5) When will the A2091 actually be available?
6) How many SCSI devices can the A2091 drive?
7) Can the A2091 be cabled to a drive in the 5-1/4" bay? To a drive in
en external chassis?
Thank you!
--
Rich Carreiro - Most Biased Boston Celtics Fan! Get well soon, Johnny Most!
ARPA: rlcarr@space.mit.edu Welcome back, Larry!
UUCP: ...!mit-eddie!space.mit.edu!rlcarr McHale, 6th man supreme!
BITNET: rlcarr@space.mit.edu "Follow through!"rlcarr@athena.mit.edu (Rich Carreiro) (11/22/89)
[following up my own article...] >7) Can the A2091 be cabled to a drive in the 5-1/4" bay? To a drive in > en external chassis? 8) Can either benefit from future upgrades to an '020 or '030 CPU? Which is helped more? 9) Will Amix be able to work with either (neither?) of them? >Thank you! -- Rich Carreiro - Most Biased Boston Celtics Fan! Get well soon, Johnny Most! ARPA: rlcarr@space.mit.edu Welcome back, Larry! UUCP: ...!mit-eddie!space.mit.edu!rlcarr McHale, 6th man supreme! BITNET: rlcarr@space.mit.edu "Follow through!"
daveh@cbmvax.UUCP (Dave Haynie) (11/23/89)
in article <1989Nov21.224423.3829@athena.mit.edu>, rlcarr@athena.mit.edu (Rich Carreiro) says: > Keywords: SuQuest HardFrame A2091 > 1) Which is faster, HardFrame or A2091? Don't know if anyone's done a comparison yet. They should be about even I'd think; should be an interesting comparison at best. They sound like they do pretty much the same kind of thing. The main places for a difference will be in FIFO size (which determine how many times you need to get on and off the bus) and the efficiency of the device driver (both are written in assembler, I believe, and both will run from the actual autoboot ROM, so there shouldn't be much foot dragging here). > 2) Does the A2091 fix the DMA/overscan slowdown problem of the A2090(a)? Completely. Obviously, there's some degree of unavoidable slowdown if you're using all the chip bus bandwidth for video display as covered in previous articles, but the 2091 knows how to deal with this as effectively as possible, so no FIFO overruns ever occur. > 4) Does the HardFrame work with the 6.2 motherboard? Yes. > 7) Can the A2091 be cabled to a drive in the 5-1/4" bay? To a drive in > en external chassis? Yes, it has both internal 50 pins SCSI ribbon cable and external Macintosh style 25 pin D connector. > Rich Carreiro - Most Biased Boston Celtics Fan! Get well soon, Johnny Most! -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Too much of everything is just enough
ccplumb@rose.waterloo.edu (Colin Plumb) (11/25/89)
In article <1989Nov21.224423.3829@athena.mit.edu> rlcarr@space.mit.edu (Rich Carreiro) writes: >I have narrowed my selection to either a HardFrame or an A2091 for a >controller, and will probably get a SyQuest rmovable hard drive. My controller choices, too. I don't know about the SyQuest, but I know nothing goes faster than a DMA controller (except a better DMA controller) and I want the speed. 1 MB/sec? Congratulations, all! I'll skip what Dave Haynie has answered, but mention a few things he didn't. >3) I've been told that the SyQuest takes 10-15 seconds to spin up. > Will this cause timeout problems with the HardFrame? The A2091?? > [note: I called MicroBotics about this - the person who answered was > clueless as to the answer] Someone here has a 2090 and a drive that doesn't spin up fast enough. If it doesn't boot, C-A-A a second time gets it going fine. Presumably, the 2091 won't be any less friendly. If you need it to boot unnattended after a power failure or the like, more testing may be necessary. >5) When will the A2091 actually be available? I'm hearing reports they'll be available at World of Commodore next weekend (near Dec. 1). *Real* soon now. If someone at C-A who knows could mention it when you start shipping the things to dealers, I'm sure it would be widely appreciated. >6) How many SCSI devices can the A2091 drive? Like all SCSI devices (unless they have really $#!+ device driver software), you can have up to 8 "things" on the bus, including the controller. -- -Colin
rchampe@hubcap.clemson.edu (Richard Champeaux) (11/25/89)
In article <18611@watdragon.waterloo.edu>, ccplumb@rose.waterloo.edu (Colin Plumb) writes: > In article <1989Nov21.224423.3829@athena.mit.edu> rlcarr@space.mit.edu (Rich Carreiro) writes: > >I have narrowed my selection to either a HardFrame or an A2091 for a > >controller, and will probably get a SyQuest rmovable hard drive. > > >3) I've been told that the SyQuest takes 10-15 seconds to spin up. > > Will this cause timeout problems with the HardFrame? The A2091?? > > [note: I called MicroBotics about this - the person who answered was > > clueless as to the answer] > > Someone here has a 2090 and a drive that doesn't spin up fast enough. > If it doesn't boot, C-A-A a second time gets it going fine. Presumably, > the 2091 won't be any less friendly. If you need it to boot unnattended > after a power failure or the like, more testing may be necessary. > I have a HardFrame and a Segate 296N. The 296N takes 20 - 25 seconds before it acknowledges any requests. At least the access LED provided by the HardFrame just barely filckers durring the time the 296N is going through it's self test. The HardFrame, however, has no problem with this long delay. My system boots without any problem from a cold start. The only difference is that it takes 20 - 25 seconds longer. Rich Champeaux (rchampe@hubcap.clemson.edu)
lphillips@lpami.wimsey.bc.ca (Larry Phillips) (11/28/89)
In <24531@cup.portal.com>, FelineGrace@cup.portal.com (Dana B Bourgeois) writes: >I know it probably isn't a problem now but when will support for more >than 8 SCSI devices appear? And do the current products support the >attach/detach elements that are built into the standard? SCSI has >been around for at least 5 years now and it seems there is only support >for a subset of the standard. What's the reason? Support for more than 8 devices will never appear in 'SCSI'. Perhaps in some future standard, perhaps in an extension of SCSI, but not in SCSI itself. Note that devices are controllers, and you are only allowed 8 of them, including host adapters in the count. Each controller may have more than one thing attached to it. For example, an Adaptec 4000 will allow two ST506 drives to be attached, and you can have 7 Adaptec 4000s on the buf, for a total of 14 drives. This is still, however, 8 devices.. 7 Adaptecs and the host adapter. Other controllers allow up to 7 LUNs (logical units) to be attached. We have the support now, in the Amiga, though not on all host adapters. -larry -- " All I ask of my body is that it carry around my head." - Thomas Alva Edison - +-----------------------------------------------------------------------+ | // Larry Phillips | | \X/ lphillips@lpami.wimsey.bc.ca -or- uunet!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 -or- 76703.4322@compuserve.com | +-----------------------------------------------------------------------+
FelineGrace@cup.portal.com (Dana B Bourgeois) (11/28/89)
[line eater food] I know it probably isn't a problem now but when will support for more than 8 SCSI devices appear? And do the current products support the attach/detach elements that are built into the standard? SCSI has been around for at least 5 years now and it seems there is only support for a subset of the standard. What's the reason? Dana Bourgeois @ Cup.portal.com And why don't we have SCSI modems? Don't tell me a 9600 BPS modem wouldn't benefit from SCSI. Better yet, why wouldn't it be a good idea?
kent@swrinde.nde.swri.edu (Kent D. Polk) (11/29/89)
In article <24531@cup.portal.com> FelineGrace@cup.portal.com (Dana B Bourgeois) writes: >I know it probably isn't a problem now but when will support for more >than 8 SCSI devices appear? [...] >And why don't we have SCSI modems? Don't tell me a 9600 BPS modem >wouldn't benefit from SCSI. Better yet, why wouldn't it be a good idea? How about ... a SCSI Tape Backup! How about a Mac SCSI scanner? Is there a hardware limitation with the 2090A or other available SCSI controllers for Amigas which will not let them communicate with such devices or just a lack of software support? ==================================================================== Kent Polk - Southwest Research Institute - kent@swrinde.nde.swri.edu Motto : "Anything worth doing is worth overdoing" ====================================================================
lphillips@lpami.wimsey.bc.ca (Larry Phillips) (11/29/89)
In <24278@swrinde.nde.swri.edu>, kent@swrinde.nde.swri.edu (Kent D. Polk) writes: >How about ... a SCSI Tape Backup! > >How about a Mac SCSI scanner? > >Is there a hardware limitation with the 2090A or other available SCSI >controllers for Amigas which will not let them communicate with such >devices or just a lack of software support? Lack of software support only. -- " All I ask of my body is that it carry around my head." - Thomas Alva Edison - +-----------------------------------------------------------------------+ | // Larry Phillips | | \X/ lphillips@lpami.wimsey.bc.ca -or- uunet!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 -or- 76703.4322@compuserve.com | +-----------------------------------------------------------------------+
a186@mindlink.UUCP (Harvey Taylor) (11/29/89)
In Msg-ID: <24278@swrinde.nde.swri.edu>, kent@swrinde.nde.swri.edu (Kent D. Polk) writes: |In article <24531@cup.portal.com> FelineGrace@cup.portal.com |(Dana B Bourgeois) writes: |>And why don't we have SCSI modems? Don't tell me a 9600 BPS modem |>wouldn't benefit from SCSI. Better yet, why wouldn't it be a good idea? | | How about ... a SCSI Tape Backup! | How about a Mac SCSI scanner? | Is there a hardware limitation with the 2090A or other available | SCSI controllers for Amigas which will not let them communicate with | such devices or just a lack of software support? | Time to talk up SCSI Direct (aka scsidisk.i|h in the includes/devices)! Wherein it is writ: ; SCSI Command ; Several Amiga SCSI controller manufacturers are converging on ; standard ways to talk to their controllers. This include ; file describes an exec-device command (e.g. for hddisk.device) ; that can be used to issue SCSI commands I think it is important that all developers/manufacturers support this command. The Amiga deserves to have a well defined, supported standard method of using any SCSI device with any SCSI controller. I know that there are difficulties in that not all SCSI devices are created equal, some will support extensions that others don't. Okay. Let the hacker community worry about it. I am sure that there are just oodles of guys out there who would love to tackle getting a CD-ROM talking to their Amiga. And if we end up with 15 or 20 drivers (Hitachiscsi.device, NECscsi.device, AMDEKscsi.device) so be it. The point is that until SCSI Direct is supported there is no standard way to even get to the SCSI bus. As a buyer you can do your bit by, the next time you are buying a SCSI controller, simply asking, "Does it support SCSI Direct?" Hey CBM, do the A590 & the A2091 support SCSI Direct? Hey C Ltd, do you support SCSI Direct? Hey MicroBotics, Hey GVP, Hey ... <-Harvey "The more I think about history, ancient or modern, the more ironical human affairs seem."-Tacitus Harvey Taylor Meta Media Productions uunet!van-bc!rsoft!mindlink!Harvey_Taylor a186@mindlink.UUCP
ccplumb@rose.waterloo.edu (Colin Plumb) (11/29/89)
A few more notes: The A590 has a DIP switch to select slow booting, to let disk drives spin up. Presumably, the 2091 will have this as well. In article <24531@cup.portal.com> FelineGrace@cup.portal.com (Dana B Bourgeois) writes: >I know it probably isn't a problem now but when will support for more >than 8 SCSI devices appear? And do the current products support the >attach/detach elements that are built into the standard? SCSI has >been around for at least 5 years now and it seems there is only support >for a subset of the standard. What's the reason? There is support for more than 8 SCSI devices, in some sense... each bus interface can have 8 "logical units" attached to it. The A590, at least, supports multiple LUNs. While not as bad as RS232, there are an infinite number of bad SCSI implementations, and some drives answer multiple LUNs (even though they're the same thing), so the feature can be disabled by a dip switch. As for why people keep supporting substandard SCSI implementations... well, becasue they work. People can plug them in and still have it work, so they don't complain. And it saved the manufacturer a few months of ROM debugging (and possibly bigger ROMs), so he's happy. >And why don't we have SCSI modems? Don't tell me a 9600 BPS modem >wouldn't benefit from SCSI. Better yet, why wouldn't it be a good idea? I believe they exist: I heard of some extra serial ports for Suns that hook onto the SCSI bus. I was talking with onr of the local gurus, and he was lamenting that we have this nice thin coaxial cable coming from the terminal concentrator, which turns into hundreds of RS232 lines, and then all get plugged into cards. The SCSI bus seems like a good thing to use for hooking up 50 terminals or so. -- -Colin
cmcmanis%pepper@Sun.COM (Chuck McManis) (11/29/89)
In article <24531@cup.portal.com> (Dana B Bourgeois) writes: >I know it probably isn't a problem now but when will support for more >than 8 SCSI devices appear? Hmm, in my copy of the spec 8 is the limit. Now the HardDrive interface will support 8/Card and multiple cards, is that what you want? > And do the current products support the attach/detach elements that > are built into the standard? Again the RDPrep routine for the Hardframe asks if this device supports disconnect/reconnect but I don't know if the driver actually understands it. --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@Eng.Sun.COM These opinions are my own and no one elses, but you knew that didn't you. "If it didn't have bones in it, it wouldn't be crunchy now would it?!"
jwright@atanasoff.cs.iastate.edu (Jim Wright) (11/29/89)
cmcmanis@sun.UUCP (Chuck McManis) writes: | In article <24531@cup.portal.com> (Dana B Bourgeois) writes: | >I know it probably isn't a problem now but when will support for more | >than 8 SCSI devices appear? | Hmm, in my copy of the spec 8 is the limit. Now the HardDrive interface | will support 8/Card and multiple cards, is that what you want? I seem to remember something about 8 devices on the SCSI bus, with up to 8 units associated with each device. Anybody know the truth about this? -- Jim Wright jwright@atanasoff.cs.iastate.edu
doug@xdos.UUCP (Doug Merritt) (12/01/89)
In article <24278@swrinde.nde.swri.edu> kent@swrinde.UUCP (Kent D. Polk) writes: >How about ... a SCSI Tape Backup! >How about a Mac SCSI scanner? How about a SCSI remote file system for linking different systems? Seems much more cost efficient than Ethernet. >Is there a hardware limitation with the 2090A or other available SCSI >controllers for Amigas which will not let them communicate with such >devices or just a lack of software support? Someone else will likely comment about hardware; I've heard some controllers support only a subset of SCSI, but myself, I dunno the issues. But there certainly are problems with software support. E.g. you need a different kind of driver for SCSI tape, because a tape drive behaves differently than a hard drive. Same for WORMs, scanners, etc. An interesting approach to a *general* solution to this kind of problem has been designed and implemented on Suns by Rich Morin of Canta Forda Laboratory, Pacifica CA (named for his home computer, a fully loaded Sun). It is described in an article in the December 1989 issue of Sun Expert (second of two parts; first part in November is a whet-your-appetite teaser on the subject). I talked to Rich about it a few months ago. The idea is to create a really *generic* SCSI driver. On unix this is of course /dev/scsi. Using this via ioctl() system calls, potentially you can talk to any kind of SCSI device at all from an application program. I don't know the details yet (haven't seen the Dec. Sun Expert), but I believe this means that you still need to write new C code for new types of SCSI devices, but that much of the complications are shifted out of the custom code and into the generic SCSI driver. Sounds like a good idea, and one that should be adapted for the Amiga. Doug -- Doug Merritt {pyramid,apple}!xdos!doug Member, Crusaders for a Better Tomorrow Professional Wildeyed Visionary
jms@tardis.Tymnet.COM (Joe Smith) (12/01/89)
In article <2037@atanasoff.cs.iastate.edu> jwright@atanasoff.cs.iastate.edu (Jim Wright) writes: >I seem to remember something about 8 devices on the SCSI bus, with up to >8 units associated with each device. Anybody know the truth about this? The specifications for SCSI is in ANSI X3.131. I have the 1986 edition. Yes, there is a limit of 8 controllers on the bus. It has to do with the way arbitration is done on the 8-bit data bus. When the bus is idle, any one of the 8 controllers may request the bus. If controller #0 wants the bus, it pulls down on the wire for bit 0, controller #1 uses the wire for bit 1, etc, up to controller #7 using bit 7. A single SCSI controller may have several logical units attached to it. Each SCSI command has a 3 bit field that can be used by the controller to address a particular unit on that controller. This means up to 8 disk units can be attached to a single disk controller. Since you have to have one controller attached to the CPU, that means you can have 7 disk controllers with 8 disks each for a total of 56 disks. Using 700 megabyte CDC Wren VI disks, that's a total of 39 gigabytes! -- Joe Smith (408)922-6220 | SMTP: JMS@F74.TYMNET.COM or jms@gemini.tymnet.com BT Tymnet Net Tech Serv | UUCP: ...!{ames,pyramid}!oliveb!tymix!tardis!jms PO Box 49019, MS-D21 | PDP-10 support: My car's license plate is "POPJ P," San Jose, CA 95161-9019 | humorous dislaimer: "My Amiga speaks for me."
FelineGrace@cup.portal.com (Dana B Bourgeois) (12/02/89)
[line eater food]
I stand corrected: a maximum of 8 devices with 8 attached logical units
on each device. What we need is for SCSI drives to be available both as
devices and 'logical units'. Is it too hard to do that? Make it switch
selectable perhaps so your second drive can plug into your first as a
logical unit instead of taking a second device address. While a moot
point now this will quickly become a problem as SCSI takes off and
equipment is networked over the SCSI bus. ("who needs more than 640K?")
Lets not be short sighted about this.("why would anyone want more than
two serial ports?")
9600 BPS modems are not standardized yet and the price is still hovering
just under $1000 but they are coming and I predict they will have a very
very (VERY) large impact. They will sell like hotcakes and the price
will drop quickly. I saw a recent ad for a Hayes 9600 BPS modem that
supports just about every standard including the error correcting ones
like MNP. It lists for around $1000 but is available to BBS owners for
about $600!! Remeber when a Hayes Smartmodem - the 1200 BPS one - sold
for $600? That was only four years ago! My prediction is two years or
less. That was the reason for my comment about a SCSI modem.
Who makes SCSI units for the A1000 and which one is the best? Is it DMA?
Is it Supra who is working on a shared file system for SCSI or someone
else? I want to be able to hook my A1000 and my next machine on the same
SCSI bus and share the peripherals. When is that gonna be reality?
Dana @ Cup.portal.com
And what is going on with 'Amiga Transactor'? They don't answer the
phone. The last issue is October. Are they in trouble again?doug@xdos.UUCP (Doug Merritt) (12/02/89)
In article <840@tardis.Tymnet.COM> jms@tardis.Tymnet.COM (Joe Smith) writes: >Using 700 megabyte CDC Wren VI disks, that's a total of 39 gigabytes! What, do you have your own power substation at your house to power these 56 mondo drives? :-) Doug P.S. Ever feel like 15 amp circuits just weren't enough? -- Doug Merritt {pyramid,apple}!xdos!doug Member, Crusaders for a Better Tomorrow Professional Wildeyed Visionary
keithh@atreus.uucp (Keith Hanlan) (12/05/89)
In article <18611@watdragon.waterloo.edu> ccplumb@rose.waterloo.edu (Colin Plumb) writes: >In article <1989Nov21.224423.3829@athena.mit.edu> rlcarr@space.mit.edu (Rich Carreiro) writes: >>I have narrowed my selection to either a HardFrame or an A2091 for a >>controller, and will probably get a SyQuest rmovable hard drive. > >My controller choices, too. and mine ... :-) What will be necessary to use the SyQuest removeable media drive with my HardFrame? With the GVP controllers, you need to add special EPROMS. I anything like this required for the HardFrame? (Sort of a long-term question...) >>5) When will the A2091 actually be available? > >I'm hearing reports they'll be available at World of Commodore next weekend >(near Dec. 1). *Real* soon now. They weren't available. :-( I did get assurances that they would be available in Canada within a couple of weeks. Too late - I wanted to take advantage of the bargains. Keith Hanlan Bell-Northern Research, Ottawa, Canada 613-765-4645 uunet!utgpu!bnr-vpa!bnr-fos!bmers58!atreus!keithh or keithh@bnr.ca Keith Hanlan Bell-Northern Research, Ottawa, Canada 613-765-4645 uunet!utgpu!bnr-vpa!bnr-fos!bmers58!atreus!keithh or keithh@bnr.ca