Adaptec 1502 - SCSI Hard Drive Question

see below, plz. Dick ----- Original Message ----- From: "Jerome Fine" <jhfine(a)idirect.comTo: <classiccmp(a)classiccmp.org>rg>; "Richard Erlacher" <edick(a)idcomm.comSent: Saturday, January 12, 2002 12:22 PM Subject: Re: Adaptec 1502 - SCSI Hard Drive Question I ever That's what's troubled me all along. The IOMEGA attempt at software has often worked ... sort-of ... but also left holes which they never seem to endeavor to fill. The 1502, BTW, is normally associated with low-performance devices, e.g. scanners, which don't demand terribly fast transfer rates for optimal operation. I'd suggest you try a faster adapter. You may find it works a lot better. Back in '90 or so when we finally got the couple of dozen '501's we had working such that all the drives would read one another's writing, etc, and the performance was reasonable by the performance standards of the day, which were not what we expect these days, BTW, it was with ISA-based 8-bit SCSI controllers that they topped out. We had none of the "high-performance" adapters, e.g. AHA 1542 available for experiments such as that. I suspect they're capable of pretty good performance with the latest drive firmware and a modern adapter. Under WIndows, that sort of operation is perfectly reasonable, and, in general, thats what I'd expect to do under any circumstances with SCSI removable-media drives. IOMEGA, however, in their infinite "wisdom" uses that SCSICFG program (or whatever it's called) that leaves a file behind that characterizes the SCSI subsystem as it was when the utility was run. If you change the configuration, i.e, if you turn off a device so it can't identify itself, the IOMEGA software won't like that and often completely refuses to boot. That doesn't mean the system won't boot, but it won't have that SCSI chain on it. There's debate about that, but under RT-11 you don't have the risk that some Windows feature will P*SS on some part of your platter. There's still debate about whether it's riskier to power a subsystem up and down as needed, as opposed to leaving it "on" all the time. However, I prefer the former, simply because the SCSI is inteded to be used that way. not and having These are both excellent adapters, but I prefer the AHA2940AU, though I've had some lifetime and support issues with them. They seem to work about as well as any adapter I've ever used, and the Windows plug-n-play stuff works quite well with it. That's where I'd look for the problem. If you use the ADAPTEC low-level driver (ASPInDOS.SYS) for DOS or sparrow.mpd as installed by WIndows, you should have no trouble with the IOMEGA interface. Unfortunately, if you don't use their mid-level driver, SCSIDRVR.SYS (??? that's what it was, IIRC, when I started using my JAZ drive) you can't use their utilities either but that shouldn't be a problem until you need to tweak some feature on the IOMEGA drive. If you can use the AHA2940, it should give you no trouble with Windows. I can't vouch for it under the DEC OS emulator. devices. Your adapter should be at ID=7. This is the convention with either narrow or WIDE SCSI channels. I'd put no drives at ID=0 or ID=1 unless you're considering them for boot. You can, of course, assuming you've got the AHA2940 in place, configure the card so it doesn't attempt to boot from any device, and you can probably also tell the motherboard BIOS to disable SCSI boot. If you want to boot from the SCSI, which I'd certainly consider as an alternative to running Windows all the time, you might want to put a power switch on the default boot drive and set the boot polling order such that it looks for the EIDE drive and, if it's unavailable, uses the SCSI to boot. That way you can boot from whatever OS you select on your bootable SCSI drive. I've used removable drives in trays for a number ofyears now, and can't imagine not having them. I even modified a tray so I can put my JAZ drive in a slot in the box, thereby allowing me to boot whatever OS is on the platter I insert, if I first disable the EIDE drive, which is trivial if I simply turn off the power to it. (The frame has a keyswitch that disables power to the drive). Because it's in a tray, I can pop out the JAZ drive and put in a different SCSI drive. The drive ID is assigned by the frame, so whichever drive is in the tray gets that ID. You can selectively configure your system by putting different OS' on different platters and putting them in the bootable drive, if you have one configured for boot. You may be pleased to find that the operation is faster with the AHA2940AU. any The a device-manager will work I background. Because I have several machines, all equipped with the cheapie ~$15 frame/tray arrangements, I can put the IOMEGA drives into a system that is running DOS and Win3.11. There I have pretty good control over what the IOMEGA drivers run off and do. I load the low-level driver (ASPInDOS.SYS) from ADAPTEC, the mid-level driver from IOMEGA (SCSIDRVR.SYS) and run SCSICFG.EXE in the AUTOEXEC.BAT (as the default, set up by their install software ... an awkward step, BTW, that doesn't always work). And that allows me to use their SCSIUTIL program to format, etc, the platters. > > > I would prefer to not use W95, but that is all I have to use when I post > > > to classiccmp and use email and the internet. > > You'll have better luck if you purge all IOMEGA software from your system. > > I've found that IOMEGA drives work better with the ADAPTEC drivers, though you if > > you boot a version of DOS that you've set up to support the IOMEGA tools > > whenever you want to perform those functions. > As I have said, I now rarely use the Iomega Zip drive, but the one thing that drive > enough to justify all that bother for that one feature. I would just > UNPROTECT all the Zip media I have right now (only about 10 at the most) > and transfer everything to the Sony media just in case. I've completely given up on ZIP drives and use LS120's (significantly slower, but I still can use the first one I ever got, while the IOMEGA ZIPs have died 8-9 times) which also serve as a standard floppy drive. > > > Plus, I really use the PC to run RT-11 on the PC, so please don't suggest drives I > > > have to. I don't even know enough about MAC's to be dangerous, so that's not likely ... > > > So, it sounds like a termination problem on the SCSI bus. But > > > the ST11200N is definitely terminated from the drive. I checked at: > > > http://www.seagate.com/support/disc/scsi/st11200n.html > > > and J2 has both the Parity and Terminator jumpers. But when even just > > > one ST11200N drive is cabled on and powered on by itself, booting W95 > > > does not recognize the ST11200N, but having the Sony drive there first > > > does the job as stated above. My external JAZ drive (I have one external, and one tray-mounted internal drive.) is in a package VERY similar to what I've seen on external ZIP drives. It has a termination switch on the back. That should be all the termination you need, provided the ZIP is the LAST drive in the physical chain. The address, BTW, doesn't matter, so long as it's unique on that SCSI chain. > > > So I am starting to think the problem is more likely the device drivers. > > You'll find that out if you try the ADAPTEC drivers and leave the ZIP drive > > off the chain for the moment. > That seems like a good suggestion. Can you e-mail me the Adaptec driver > for the 1502 host adapter I am using? Plus a few helpful hints? Please > realize I am a PC dummy - I just don't want to spend the time to learn > Windows. I am 63 now and I know RT-11 very well, but that is where > I want to spend my time. So I will likely continue to use the PC only for > e-mail (internet access in general) and especially to run RT-11 under the > E11 emulator. If you don' thave the ADAPTEC drivers for DOS, I'll happily send 'em to you. You won't need 'em under Windows, though, and I doubt they'll be of much use under that RT-11 emulator unless it runs under DOS. > > > Could the 1502 and the Adaptec device drivers just not recognize an > > > ST11200N (or any other SCSI hard drive that is not removable), but > > > be fooled once booted with the Sony drive which has a removable media? > > I've found the 1510/152x controllers work about as well as SCSI adapters can > > be expected to work, and with a WIDE range of devices. I used the SONY MO > > drives with them for some time, until they became too valuable to keep. > Skip the following if you are not interested in what I did 10 years ago! > Right now I have the Sony SMO S-501 drives which are the original that > Sony released about ten years ago. In fact, that is how I started. At the > time, a customer NEEDED a better way of backup for his RX02 8" floppy > disks. There were about 1500 floppies using an RL02 as the backup - > 20 * 8" floppies of 1/2 Mbyte each on a 10 Mbyte RL02 for a total of > about 75 RL02 disks - or about 3 very large cabinets. Since there were > 5 natural groups of about 15 RL02 disk packs each, the Sony media with > 295 MBytes per side or about 8.75 RT-11 partitions was VERY sufficient > to hold those 15 * RL02 disks worth of files on 5 RT-11 partitions. > This 5 sides of media on 3 disks was then sufficient to hold the same > information as 75 RL02 disks. And then there was the REQUIREMENT > for 2 off-site backups. Originally, the 75 RL02 disks each had two tape > backups off-site. Or about two more cabinets of storage. Now, all NINE > MO media (one primary copy and 2 backup copies) were the total size > of a loaf of bread and could be held in one small box less than the total > volume of one RL02 disk. > Eventually, ONE partition was needed for a VMS directory structure > when a VMS/VAX system was also required to be able to share the > backup. That hard part in that case was to generated (very carefully) > a file structure in VMS in the first 32 MBytes of the MO media (RT-11 > partition zero) seven contiguous VMS files which exactly overlaid > each of the other RT-11 partitions - not difficult in principle, but > impossible in practise since VMS does NOT allow the user to > specify where files are placed. I set up an RT-11 system right beside > the VMS/VAX system with its own Sony drive. By fooling VMS > (in RT-11 I can easily look at any block on the media and I simply > found out where the VMS free block list was located - and zeroed > out all the blocks except one 65536 block set that were at the same > location as an RT-11 partition - actually it was not simple and the > whole exercise took 3 days to figure out and do the job), I was able > to allocate 7 contiguous VMS files which could be MOUNTed in > Exchange under VMS and provide access, in turn, to each RT-11 > partition. The problem in Exchange in VMS was that if the whole > 295 MByte media was MOUNTed as a foreign volume, ONLY > RT-11 partition zero was available - there was NO parameter in > VMS Exchange to say which RT-11 partition was desired. > I see the drives on eBay and have purchased a few along with more than > 100 media that were about to be tossed. For RT-11 which still breaks > everything up into 32 MByte partitions like V3.2 of MS-DOS did, these > Sony drives and media of 295 MBytes per side (8.75 RT-11 partitions) > are really ideal. I can backup an ST32550N on just 4 media which is > probably just about all the RT-11 files I will ever have - well, it might be > double or even triple that if I count all the duplicates, but definitely not > ten times. I once believed the SONY '501's were pretty nice drives ... however, things have changed, and I really like using interchangeable drives in trays on my SCSI chains. If you have a few, say 3 or 4, identical drives in trays that allow their swap, and you can freely hot-swap them if you disconnect power (I mention this only because some of my trays require you to do that to unlock the tray, while a couple don't.) and I've found that SCSI drives cost substantially less than the equivalent amount of storage in the form of 1 GB JAZ media. (I just bought a 9.1 GB drive for $32, and plan to snag a dozen or so more if they're available at that price.) > > > NOTE: I want to use the ST32550N under both RT-11 as a permanent > > > and much faster hard drive (the Sony is quite slow) and W95 as a backup > > > device - I have an option on a bunch of these drives even if I can't solve > > > the boot problem - does anyone else want a few as well? > > What do you mean by "W95 as a backup device?" Doesn't RT-11 support backup? > OK - some more definitions. I now will use about 5 of the ST32550N > drives to perform a backup of all of the files on my 1.2 GByte ESDI hard drive adequate month. Be careful what you call a "backup," particularly under Windows. Whereas, in some environments, a true copy is an adequate backup, under Windows 9x, nothing that appears to the operating system as a drive is suitable as backup. Once Windows recognizes and registers a drive, any virus residing anywhere in the system will be automatically notified of the presence of a new environment to contaminate. What's safe is a medium, DVD-R, Tape, etc, that requires an intermediary program to effect the data transfers under its own control without the intervention of Windows. Once a drive appears under "My Computer" you can safely assume it's infected with whatever may infect anyting else in the system. > The other drives I will begin to use under RT-11 as the primary source instead the > newer W98 system with the Pentium III 800 and 30 GByte dual IDE drives > up and running in the next year or two), I will switch to using the ST32550N > drives as backup in W95 and primary plus backup in RT-11. > Can you use any ST32550N drives? They were originally in an AlphaServer > as a RAID - about 18 drives in the AlphaServer. I don't even know what ST32550N drives are. The numbering suggests a 3.5" drive with 2 platters and 550MB capacity. The 'N' seems to suggest narrow SCSI interface. If that's correct, I'd suggest you hang onto them and use them in removable trays (visit www.ucdweb.com) since they're identical. IIRC, the MO drives were single-sided, though you could use both sides of the media, so you get only 275MB per side from them, while these will be 10x as fast and about twice the capacity as well. > > > Question: Can anyone HELP? Does anyone have any experience with the > > > device drivers from Iomega for their Zip drives? If that is the problem, then the is the > > > answer, then I can probably live with that for a few more years. > > > Thanks in advance for any suggestions! I don't think recabling will be required. You should have a termination at each end of your cable, however, and the default configuration of your AHA2940 will see to it that the adapter end is properly terminated. If you have both internal and external drives, they must have a termination at the end away controller in the middle. In the position at the "middle" of the SCSI chain, the controller will not terminate the chain. Be certain that at least one device on the chain has power to the SCSI terminations (TERMPWR) enabled. You can safely enable this option on ALL devices, as they should be both fused and diode isolated from the chain. This means that only one device provides power until some other device offers up a higher TERMPWR supply. > If anyone else is reading this, perhaps you can add to the discussion. > Sincerely yours, > Jerome Fine