Questions and Answers on MFM drives from the Ontrack BBS - 10/24/90 
-------------------------------------------------------------------------
QUESTION:

Simple question:  Has anybody fixed the problem related to cyl>1024
and/or smartdriv.sys and/or smartdriv.sys with DM ?
Has anybody come up with another cacheing prog to use in place of
windows 3.0 smartdriv.sys??
I tried hyperdsk410 but still had the drive go down <scrambled fat>.
It has been back up and running windows 3.0 in enhanced but without any
cacheing and so far no scrambling. Please don't suggest contacting
Microsoft.  We see their approach:  "....you many experience problems
when using certain third party partioning software...". Then they
procede to list virturally all third party partioning software!

ANSWER:

I see that prior to asking your question you read bulletin number 12
that has to do with the Windows/DMDRVR/SMARTDRV situation.  The reason
that bulletin number 12 exists is so I don't have to go through the
whole story every time someone asks about Windows, and I won't.

There have been a few minor changes in the situation since the bulletin
was written.  Ontrack has volunteered to make a few modifications to
XBIOS.OVL that will make it unnecessary to place the "virtualHDIRQ=OFF"
line in your SYSTEM.INI file when using a disk with more than 1024
cylinders in order to prevent Windows from hanging when attempting run
it in 386 enhanced mode.  This forthcoming version of XBIOS is
Ontracks' contribution to the solution of the overall problem.  All of
the other conflicts are at this time being addressed by MicroSoft,
since it is SMARTDRV's ignorance of the disk environment that is
causing the other problems.  (I was always taught that "ignorance" is
not a bad word in and of itself.  It simply means a lack of awareness
in a specific area and is easily remedied.)  Our engineers have been in
close contact with MicroSoft and have given them all the necessary
information that they need in order to understand the disk environment
and make SMARTDRV a little smarter.  The same functionality will have
to be incorporated into Windows itself as well in order to make it
possible to create a permanent swap file when running Windows on a
nonstandard disk.  It is my understanding that MicroSoft is currently
working on these updates.

As far as other caches being available, the answer is yes.  "Super PC
Qwik" by MultiSoft is one of probably several caches that have been
shown to function correctly on a disk with greater than 1024 cylinders,
and with Windows in 386 enhanced mode.

Thanks for your questions.  It gave me the opportunity to bring the
current status of the situation out into the light for others to see as
well.

I know this is going to sound strange, but believe it or not this whole
Windows/DMDRVR/SMARTDRV situation has had a bright side.  The fact that
this conflict happened has forced open some doors between Ontrack and
MicroSoft that neither company thought would ever be opened, and that
can be nothing but beneficial to both companies, (and the public), in
the future.

QUESTION:
 
Thanks for the update on the Windows\Smartdriv issue and that PCKwik is 
ok to use as a replacement. Please note:  the version of PCKwik to use 
may be critical.  I read/heard somewhere the version is critical 
relative to dos level and windows level. 
I found a cache (HYPERDISK) that appeared to imply it was safe with 
Windows 3.0.....but my drive still got scrambled. Since I've stopped 
using any cache the drive has stayed up.  
Do you have any comments on HYPERDISK (v:4.10), a shareware cacheing program?? 
Thanks again.
 
ANSWER:
 
It is true that you need to be careful of which revision level of 
Multisoft's Super PC-Kwik you use.  Early versions of this product did 
NOT support DM partitions or a disk with greater than 1024 cylinders. 
I believe the minimum revision level of this product needed to support 
a disk with greater than 1024 cylinders is their v3.251, (which I also 
believe is rather old by now).  I have no idea as to which version is 
compatible with Windows. 
 
In general, when I suggest a certain software package, I am always 
speaking of the most current version available.  Specific version 
information is best gotten from the manufacturer of that particular 
software package.  Obviously, they should know more about their 
products' capabilities than I.  Multisoft's phone # is: (503)644-5644. 
 
I have never heard of a package called HYPERDISK, which means that I 
seriously doubt if it is compatible with a disk having greater than 
1024 cylinders.  While it may have been compatible with Windows, your 
scrambled FAT indicates that it must not have been compatible with this 
size of disk. 
-------------------------------------------------------------------------
QUESTION:

I HAVE A 286 CLONE MOTHERBOARD, A SEAGATE ST225 20M DRIVE, A DATA 
TECHNOLOGY 7280 1:1 INTERLEAVE CONTROLLER, AND I'M TRYING TO RUN COMPAQ 
3.31 DOS.  I'VE STARTED DISK MANAGER WITH THE DM/4/M FLAGS, AND I'VE 
BEEN ABLE TO INITIALIZE THE DRIVE JUST FINE.  BY THE WAY, I'M USING CMOS 
DRIVE TYPE 2 FOR THE HARD DISK.  MY PROBLEM IS WITH THE PARTITIONING. 
I'LL GO THROUGHT THE PARTITIONING PROCESS, STARTING WITH DEFINING NEW 
PARTITIONS (N), THEN I CHOOSE TO DEFINE THE PARTITIONS BY MEGABYTES.  I 
CHOOSE THE DEFAULT CHOICES OF A 2M DOS PARTITION AND A 19M READ/WRITE 
PARTITION.  WHEN I TRY TO SAVE THE PARTITION TABLE AND GO INTO THE 
PREPARATION MENU, I GET BOUNCED BACK TO THE MAIN MENU WITH THE NO VALID 
PARTITION TABLE MESSAGE.  WHAT'S GOING ON? 

ANSWER:

If you cannot write the partition map to the disk successfully, it 
means that somewhere along the data path things are breaking down.  The 
data is never getting to the drive, or the drive is broken. 
 
Most often this problem is due to miscableing the drive.  Double check 
that you have the cables connected to the correct connectors on both 
the drive and controller card, (especially the narrow cable on the 
controller end being connected to the correct connector), and that they 
are not on upside down.  Otherwise, you may have a bad cable, 
controller, or drive. 
-------------------------------------------------------------------------
QUESTION:

I CAN"T GET DMN V3.11 MODGEN TO WORK WITH ADVANCED NETWARE V2.15 REV C.
NONE OF THE NOVEL DISKS HAVE A FILE CALLED VBIOSOVL.OBJ OR ANYTHING CLOSE 
TO IT. 
 
ANSWER:
 
DMN v3.11 works just fine with Advanced NetWare 286 v2.15 Rev C. 
You are correct in saying that the file VBIOSOVL.OBJ does not exist on 
any of your NetWare diskettes though, it resides on the DMN diskette. 
When prompted for it, insert the DMN diskette. 
-------------------------------------------------------------------------
QUESTION:
 
I WAS WONDERING ABOUT THE COMPATIBILITY OF THE SEAGATE DRIVES 
(SPECIFCIALLY THE 251 AND "DISK TECHNICIAN."  I KNOW SEAGATE DOES NOT 
RECOMMEND USING ANY OTHER DISK UTILTIES?  CAN I USE DISK TECHNICIAN TO 
PERFORM A LOW LEVEL NON DESTRUCTIVE FORMAT? 
 
ANSWER:
 
According to my information, Disk Technician is not compatible with a 
disk that is "nonstandard" to the system, or has a partition that is 
greater than 32 Meg in size, or has greater than 31 sectors per track, 
or has more than 1024 cylinders, or is using a translating or caching 
controller. 
 
The first two things may have a bearing on your situation.  If your 
computer has a CMOS drive type that is 820x6x17, and you have 
partitioned this drive with FDISK, then Disk Technician should be OK. 
Otherwise, I would not recommend using it. 
 
If you are looking for a comprehensive disk diagnostic package that 
would be able to do a nondestructive low-level format on this disk for 
you, (among many other things), I would suggest calling our sales 
department, (or your favorite distributor), and ordering our package 
called "Dosutils". 
-------------------------------------------------------------------------
QUESTION:
 
Hi... This is the second time calling and find the system very 
interesting... I have the following problem that perhaps you can help me 
with... 
    System...IBM/XT Clone 
    2nd HD...Microscience HH-1090 
    HD Controller...WDXT-GEN2 PLUS 
Problem... I can not format the 2nd hd (Microscience HH-1090) for 80 
megs. It only formats to 20 megs. Wondering what I'm doing wrong. I'm 
using a Disk Manager that has Microscience hd on it but not my model. 
Even changing ,or attempting to change) the drive specifications in the 
program to that of the HH-1090, still I get only a 20 meg-614 cyl 
format. Can you tell me what I am doing wrong??? Thanx in advance 
 
ANSWER:
 
An XT style machine can not use a drive with more than 1024 cylinders. 
Since the Microscience HH-1090 has 1314 cylinders, the absolute most 
you are going to get out of this drive when installed in this machine 
is 59.5 meg. 
 
(1024 cyls x 7 heads x 17 spt x 512 bps = 62,390,272 bytes = 59.5 meg) 
 
What you will need to do in order to get 59.5 meg out of it is to low- 
level format the drive through the controller's onboard BIOS using 
debug, and "dynamically" configure the drive.  Then use DM to partition 
the drive only.  Bulletin number 4 on this BBS talks about low-level 
formatting with your XT controller's BIOS.  Read or download that 
bulletin, and if you have further questions about that procedure, you 
should give Western Digital a call.  Their tech support number for 
support on their controller cards is: (800)777-4787.  Once you have the 
drive "dynamically" configured and low-level formatted, you should run 
DM in manual mode, (DM/M), and go DIRECTLY to the partitioning menu 
where you will create and prepare your partitions. 
-------------------------------------------------------------------------
QUESTION:
 
My company has purchased a set of your Superproms for use in an IBM 
AT.  After installation the BIOS would not pass control to the Super- 
Proms, so I entered DEBUG and looked for the Superproms per the proced- 
ure in the documentation.  I did find them, although they were at 
address E800:0 instead of E000:0.  I am concluding that I need a BIOS 
replacement.  The current BIOS is dated 1-10-84.  My question is: 
What replacement do I order and from whom? 
 
Is there a specific Version of the BIOS or a certain date after which 
they are all compatible?  Do I get them from IBM or from you or from 
someone else?  I would appreciate any help you can give. 
 
BTW, we are using the Superproms as opposed to DM because we want to 
install OS/2 on the machine.  We have already installed an Intel Inboard 
386.  Can you verify compatibility between the Superproms and the 
Inboard386? 
 
ANSWER:
 
Although the date 1/10/84 usually sends shivers down my spine, in this 
case I don't think it's your BIOS that is at fault.  I think it's 
probably a conflict with the Inboard.  My database states: "Some AT 
accelerator boards prevent Superproms from functioning because 
controller never gets passed to them."  Whether the Inboard 386 is 
included in "some", I don't know.  What happens if you pull out the 
Inboard? 
 
It's my understanding that the sockets that the Superproms are supposed 
to be placed in are hard coded at E000:0.  The fact that you saw them 
at E800:0 leaves me scratching my head.  I've got a couple of theories 
though...  It could be that you have the wrong size Superproms for this 
machine.  (I believe that the early ATs had 128k ROMs and the ones 
since then had 256k ROMs.  You can find out for sure by identifying the 
size of your system BIOS ROMs and then making sure you've got the same 
size Superproms.)  Or, did you by chance remove the BASIC ROMs from 
your machine and put the Superproms in their place? 
 
This is an interesting one.  Try the things above and let me know what 
you find out if you would. 
 
QUESTION:
 
The machine I was working on is being used rather heavily, so I haven't 
been able to get to it to try anything else.  However, I have another AT 
with the same BIOS date (1-10-84) that does not have an Inboard 
installed.  Guess what!  Same symptoms!  Control is not passed to the 
Superproms and they are located at E800:0 not E000:0.  BTW, your 
documentation states "The SUPERPROMS do not claim a memory address 
themselves; they allow the computer to assign them the memory location 
that they will occupy.  They will normally be given the E000 memory 
page, but this can vary from system to system." 
In any case, control is not being passed to the Superproms. 
There does not seem to be any address conflict.  Quarterdeck's Manifes 
Manifest shows BC00-EFFF as unused (with the superproms removed).  I did 
NOT remove ANY ROMS, PROMS, RAMS or anything else, I only added the 
Superproms.  The only remaining possiblities that have been mentioned is 
that the Superproms are the wrong size for the BIOS or the the BIOS is 
too old.  As far as the size, the machine that I am now experimenting 
with has the number TMM23256P-5878 and -5879. 256K right?  The other 
machine has some other number (I can't get to it right now) that 
contains neither a 256 or a 128.  Is there some way that I can tell the 
size of the BIOS through DEBUG or some other utility?  BTW, the third 
byte of the Superproms is 40H.  That means it is in fact a 256K-bit 
chip, right? 
Anyway, since all the other possibilities seem to be eliminated, that 
still leaves the BIOS being too old as the apparent problem.  So, If 
that is true, where can I get a BIOS upgrade for that machine and 
exactly what do I order? 
 
One other thing,  I had, several months ago, installed both Superproms 
and an Inboard in a newer AT and although we had some other problems 
apparently connected to the Inboard, the Superproms performed their 
function properly.  The date on that machine is 6-10-85.  That machine 
is 8MHz while the two in which the Superproms do not work are 6MHz. 
Will the 6-10-85 BIOS work in the older machines or is there a special 
upgrade for the 6MHz models. 
 
ANSWER:
 
I spoke with one of the engineers that was involved in developing 
Superproms, and I think we figured out the cause of your problem.  We 
tried a few experiments with an IBM AT and this is what we found:  If 
you place a set of 128k Superproms in an IBM AT that has 256k sockets, 
they will show up at E800:0 because of the select line being wrong for 
the chip size.  Also, we confirmed that the third byte of 128k and 512k 
Superproms is 40H, and the same byte in 256k Superproms is 80H.  It 
sounds like you actually have a set of 128k Superproms here, and would 
need to get 256's instead. 
 
While it is true that the actual code within Superproms is relocatable 
and does not need to be loaded at any particular address, the IBM AT's 
add-on ROM sockets are indeed hardcoded to E000:0, so this is where 
they should show up at. 
 
You keep asking me how to upgrade your BIOS, and I don't really know 
what to tell you.  Ontrack Computer systems does not sell BIOS ROMs. 
It's my understanding that the only way to can upgrade your BIOS 
through IBM is to send in your system board for a "fixed price repair" 
through one of their service centers.  In the process of "repairing it" 
they will upgrade the ROMs.  I believe IBM's Customer Support phone 
number is: (800)426-2468.  Another alternative is to get a set of third 
party BIOS ROMs such as Phoenix for your machine.  I have two possible 
numbers for Phoenix Technologies in my database.  One of which ought to 
at least get you in the front door.  They are: (617)769-7020 and 
(617)551-4000. 
 
CLOSING COMMENTS:
 
Funny that we should arrive at the same discovery.  I had a chance to 
work on the machine that has the WORKING Superproms installed and 
noticed that it had 80 as the third byte.  I then opened it up and 
noticed that the chips have the same stick-on labels as the non-working 
ones, but the Toshiba number is TMM27256BD-15 as opposed to 
TMM27128AD-15 on the ones that don't work. 
-------------------------------------------------------------------------
QUESTION:
 
Have been trying to get novell on a Compaq Deskpro system for over a 
week now.  Have replaced original cdc (30mb) drive with seagate st-4096 
with Western Digital Controller and then used regular disk manager (that 
came with the drive) to initalize.  Used dos 4.0 to fdisk and format. 
ran fine.  then installed novell els-I ver 2.12.  resulted in 1780 hard 
drive failure.  replaced drive.  used Compaq setup disk to set up as 
drive type 4.  Used Disk Manager to Initialize, then installed novell. 
Got Abend, Attempt to configure a non-operating drive.  Got same result 
as type 12.  Also saw a few "Problem with drive 00: error reading disk 
redirection information on sector 14.  Compsurf also refused to format 
the drive.   At suggestion of Novell distributor, ordered Disk Manager 
Novell (just registered it on BBS). Ran dmn /m and entered bad tracks, 
initialized and prepared netware partition (your manual doesn't mention 
els-I v 2.12).  When done, ran ELS Install.  Before installation 
completion, got: Abend improper rom parameter table for at hard disk 
controller.  Your manual suggests running modutils.  I tried it.  There 
is no disk with prepare.exe in els-i.  skipped it.  got aborted out when 
trying to fulfill requests for install.exe and net$os.  
 
ANSWER:
 
In order to install ELS level 1 v2.12 on a nonstandard drive, you will 
need to have v3.11 of DMN.  A call to our sales department to upgrade
to version 3.11 of DMN will solve your problems.

QUESTION:

I GOT V3.11 OF DMN WITH MODGEN.  IT RAN WITHOUT A HITCH.  INSTALLATION 
THEN RAN GREAT.  HOWEVER, WHEN RUNNING NET$OS (AFTER DMN/I) I GET ABEND:
IMPROPER ROM PARAMETER TABLE FOR AT HARD DISK CONTROLLER.  IF I TRY NET$OS 
WITH I GET ERROR READING DISK REDIRECTION DATA ON DR DRIVE 00.  
 
ANSWER:
 
Are you running the same NET$OS that you modified with MODGEN? 
If you have successfully modified this copy of NET$OS, then it is 
not possible for the "Invalid ROM Parameter Table" message to 
occur.  The two are mutually exclusive.  That's all the patch 
does is to strip out this message.  Try running MODGEN on clean 
copies of your NetWare master diskettes, then use that  modified 
copy of NET$OS to boot the server. 
 
Your message said "IF I TRY NET$OS WITH I GET ERROR READING DISK 
REDIRECTION DATA ON DR DRIVE 00."  What did you mean by this? 
Did you mean that if you don't run DMN/I before running NET$OS 
that you get this error message? 
 
Running NET$OS without running DMN/I first would indeed cause a 
"Error reading disk redirection data" message.  Since the 
nonstandard parameters weren't installed into low memory with the 
DMN/I command, NetWare would be looking for it's redir area in 
the wrong place. 
 
May I ask why you aren't trying to boot the server from the hard 
disk?  The ST4096 does not have more than 1024 cylinders, you 
don't have to boot from a floppy unless you are running 
nondedicated. 
 
P.S.  You stated that you were using a Western Digital controller 
card.  What is the model number of this card?  If it is a 
WD1002-WA2, it may not be able to support a drive with greater 
than 8 heads, depending on the date it was produced. 
-------------------------------------------------------------------------
QUESTION:
 
I have an AST 286 Premium with a Priam 45 meg hard drive partitioned 
into 32 meg and 10 meg pieces.  I am using Priam's EDVR.SYS to run the 
disk.  I have a copy of XENIX 286 I would like to install on D: (the 10 
meg part).  For Priam's manual, the XENIX installation looks involved, 
for instance, it appears that I would have to start from scratch and 
make the XENIX  partition the first one.  Is DM easier to use in this 
situation?  Will it work with Priam drives?  Can I have DOS on a 32 meg 
C: and XENIX on a 10 meg D:?       On a different subject, I read you 
TOPTEN file which said that many disk caches have trouble with DM.  If I 
stick to sectors of 512 will they work?  I do database applications that 
definitely benefit from a disk cache.   Thanks. 
 
ANSWER:
 
While it is probably possible to create a XENIX partition on your disk 
with DM, (depends on which XENIX it is, SCO system V for example would 
be "other" type 2 from within the DM partitioning menu), you would not 
be able to "prepare", (high level format), it with DM.  You would have 
to let XENIX do that itself.  You are probably best off using the XENIX 
to create it's own partition.  If the disk is nonstandard, I know that 
SCO XENIX for example will allow you to specify the parameters of the 
disk directly at install time.  Once you get the disk partitioned for 
the two OS's, DM would be handy for changing the boot partition back 
and forth between the two OS's. 
 
As for your question about disk caches, 512 byte logical sector size is 
not the only consideration.  You need to be aware wether or not your 
cache can handle a disk with more than 1024 cylinders if using that 
size of drive.  Current versions of Super PC-Kwik is one cache that I 
am aware of that will function fine under these scenarios. 
-------------------------------------------------------------------------
QUESTION:
 
Modgen ask for the file isadisk.obj.  I cannot find this file anywhere 
on the Netware disk. What to DO?? 
 
ANSWER:
 
If you are installing an older version of NetWare that uses ATDISK.OBJ 
instead of ISADISK.OBJ, you will need to select 2.1x SFT from within 
MODGEN instead of 2.1x ADV.  The SFT selection specifies ATDISK rather 
than ISADISK, and that's the only difference between the two. 
-------------------------------------------------------------------------
QUESTION:
 
When using 
DMN VER 3.10 and running MODDRVR on Netware 2.15c, I cannot get the 
proper files to modify for non-standard operation. I keep getting 
invalid file msgs and the program aborts. Any suggestions? 
 
ANSWER:
 
The "Invalid file" message means that this version of DMN does not 
understand your newer ISADISK.OBJ file.  DMN v3.10 is rather old at this 
point, and you will need to contact our sales department and upgrade it 
in order to patch this version of NetWare for a nonstandard drive 
installation.  If you have just purchased this copy of DMN, you should 
approach your distributor and inquire about getting an exchange for the 
current version. 
-------------------------------------------------------------------------
