Re: Strange PEER error with Dani's 506 1.81 generic question

Pardon the time for the reply Peter but a good reason for that ..

Peter Brown wrote:
Hi Mike

Tip: Burners seem to prefer to be Master with ROMs as Slave rather than the reverse. Thus one better be darned careful

about setting up an IDE CD device as a MASTER on any cable! But that flies in the face about data I/O speeds, as well as the requirement, as I understand things, that the CD device has to be a MASTER to be used for boot purposes if you are using, for example, the IBM install CD-ROM disks, or a modified version of this technique for custom boot CD-ROM operations.


Sorry but I'm not sure what the problem is here...

Well, first .. I lost about eight hours of research in this mess due to a complete defective batch of CD-RW new disks from Memorex. I'd run out of CR-RW disks. I bought a new 25 disk bulk pack, started using them for the research. All of a sudden, TOTAL strange failures that popped up on four different study systems and boxes that were the target of the new Dani tools! I was totally blown away by this. But it persisted even after backing the new tools from Dani out and going back to testing things with these new CR-RW disks!

I couldn't figure out what I could have done to corrupt the whole mess. Then, out of frustration, I went out and bought a box of ten new Verbatim CD-RW disks of the exact same specs I'd been using for years and were the same for the new Memorex disks. POOF! Instant success! I then re-installed all the new Dani tools for test and POOF! Instant success with her tools on these Verbatim CD-RW disks as well.

Took the whole wheel pack of Memorex disks back to Best Buy. The tech person told me they had had some other failures here as well. What they told me was that in the press process, from time to time the layering and stuff can get mangled during the manufacturing process. And, in fact, the whole batch of whatever will, indeed be bad. So they simply gave me a refund for the whole deal and I finally finished the initial research for Dani's new tools for what I have here to study with Verbatim disks.

To my experience if you are going to use the CD-ROM as a boot device, then for all of the boxes I've built from scratch, it is true that the CD-ROM device has to be the Master for the box to boot from it at all for installation purposes or even for a boot run for that, if your boot device roster proceeds from, say, Floppy Disk A:, thence to CD-ROM (Master), thence to the hard disk device, whether that hard disk is an IDE or SCSI hard drive.

In the case of everything so far up until the Intel 915GAVL/GEVL and still in research 945 series motherboards, there have been two IDE controllers on the boards. Thus the primary hard disk controller from which the finished box will normally boot would be the Master on IDE controller #1. The CD-ROM would be the Master on IDE controller #2. Thence comes the burner issue. In this case, if I additionally have a CD-RW unit, it would be the Slave on IDE controller #2 on the box. And .. if there were, for example, two IDE hard drives, one would be the normal boot device as the Master on the primary IDE controller. The other IDE hard drive would be the Slave on the primary IDE controller.

In this way, complete speed for hard disk I/O would never be restricted to whatever issues might arise for having a slower device as in the CD units on the same IDE cable as the hard drive.

In the case of the Intel 915GAVL/GEVL units, there is only one IDE controller and cable available on them. You can, indeed, boot from an IDE hard drive on these boxes. As well, you can, indeed, with at least the older Dani tools, or even the latest IBM1S506.ADD driver level, boot and run from either the IDE cable or the SATA drives. You can, indeed, clone an IDE hard drive on these Intel boxes into an SATA driver. And then, yank the IDE hard drive, and boot right up from the SATA cloned drive.

In that case, if you intend to boot from the CD-ROM drive for some reason, on the single IDE controller cable, and also use a CR-RW unit or a DVD burner, at least with the RSJ product, you'll need the burner as the Slave. Which is exactly what I have on one IDE unit in research here. As well as an older ASUS mobo box with an AMD CPU I use with mobile drive trays for research with OS/2 and Windows whatever. Which in that case has the IDE mobile drive tray on the first controller with the Master designation, and the CD-ROM as Master on the second controller and an LG DVD burner as slave on the second controller.


It has been a while since I last used the good old command line

cdrecord2 -scanbus > scsi.txt

to generate a report of scsi and "fake scsi" devices on a mixed scsi hard drives/ide dvd burner and dvd-rom system but I seem to recall that all scsi devices got listed before any ide devices - including the scsi scanner I had.

Maybe someone else can check that for you.

cdrecord2 is a package for creating and burning cds, you should be able to find the current release on hobbes.

I already have that package here, but I have never tried it. At a point a little later in this research I will do that. Particularly after studying the documentation for the latest DaniATAPI tool.

The speed issue you mention is surely related to mixing an atapi device and ide hard drive on the same ide controller - the hard drive would be as slow as the atapi device.

Nope, for sure not in at least the original failure box in this case! In this case the first controller has the IDE hard drive on it, obviously the Master drive here. It has nothing else on that IDE cable and controller. In this case, the CD-ROM, and no burner at all, is on the Master position on the secondary IDE controller on a separate cable.

As regards booting from an atapi device:

some of that is BIOS related ie if the BIOS allows booting from atapi device 2 then it will work for some boot CD/DVDs.

But I have never used the configuration of trying to boot from an atapi device like this on the same cable as an IDE (generic) hard drive.

Some software insists on a particular device because the software has been built to work that way eg Bootable http://www.xs4all.nl/~hrbaan/bootAble/index.html

I've not gotten to the research game on Bootable yet. I have made careful note of what you have taught me as well as ..

When you create a BootableCD using that software it will want to boot from whichever device it was burnt to eg I recently created a BootableCD in drive V: (dvd burner) and when I tried booting with the disc in drive W: (dvd-rom) it failed - because the BootableCD is particular to the system that it is created for and requires the same drive letter as was used during creation.

WOW! That's not at all 'good' as I see this as a novice at this.

Long time since I last installed Warp4 so cannot remember if there are any limitations with the install CD... I seem to recall that the install always completed if there was a copy of the install cd in both burner and rom (Master and Slave) but could fail if there was only a copy in the burner.

Precisely. The boot and install operation, as I've seen this for Warp 4, has to be designated as Master to even try this. And the switch disks for whatever in any install process, in this case, don't work and the whole thing doesn't work if any second CD/DVD device is actually in the Slave position on the cable.

Yes, all kinds of varients as to whether or not SCSI criss cross operation is done can be controlled through command line changes in the CONFIG.SYS.

Precisely correct as I've seen this. Including as noted here, at least for me, even a 'correct' order of driver presence in the CONFIG.SYS file for SCSI and IDE system use. I've always just set it up 'correctly' to start with, never experimented with later MCP2 changes in this order. Which has been suggested may work in the latest versions of MCP2 and so on.

But .. also, we face BIOS setup and boot preference issues where, for example, the 'normal' boot order, for setup, core level maintenance, are first device, floppy diskette, second device CD ATAPI, and third device SCSI hard disk on Drive C: in my case.

BootableCD may have the option that you need as it does not need to pretend to be drive A: to boot - uses memdisk instead.


I'm learning from you.


I think you should setup 1 box and see how it goes without "defaulting" to no scsi translation.


Are you suggesting I do this on actual SCSI system boxes which only use the IDE controller for a Master CD-ROM and a Slave CD-RW or DVD burner on the same cable? Such as for the Intel 915GAVL/GEVL actual SCSI systems that are here? As well as for one of them which has an actual PlexWriter 12/4/32 SCSI burner which is designated as an actual SCSI device on the Adaptec controller, plus has an actual Seagate DD3 tape backup drive on this, all even properly set up with spindle sync settings and so on?

Recall that if you are using spindle sync negotiation for a SCSI hard drive, you also have to have a SCSI tape drive which can be set for this as well? For exactly why the choices have been carefully made for the whole setup all over all of this to use the Seagate DAT tape drives. The can enable spindle sync negotiation as needed.

What is the likelihood of ever being able to handle this between export of any of this ATAPI stuff into a 'SCSI unit' as is contemplated in all this now under this technique I'm just discovering for CD-RW and DVD burners?

Remember that one of the worst corruption errors you can face in the BA2K Server Pro operations with SCSI tape drives, is related to long block SCSI channel writes. BA2K Server Pro is one of very few tools, so I was trained in their service support work they did with us, which uses the 64K long block write technique to the tape drives. Unless you very carefully work out the block length and choose the SCSI setup parameters for your SCSI tape backup drives you can hit HORRIBLE problems with locked systems and bad data issues here with the long block technique.

How would any CD-RW or DVD burner operation which ghosts things into SCSI ever get along in any actual SCSI based system doing this as you see it?


I think you will find it all falls into place and works well with minimum human interference.

With all the above in mind, I am sincerely thinking about your suggestion though..

As you mention RSJ you may want to have a look through those docs and you will probably get the idea that people have been using atapi burners for quite a few years with a lot of success.

It is possible (but unlikely) you may run into unforeseen problems but that is what ngs are for :-)

Have fun

Pete

Yes .. that is why I'm posting here, trying to help me as well as perhaps help others at all this!

OK .. as of now, with DaniS506.ADD and DaniATAPI.FLT, and the /!SCSI command line block, I am now apparently stable and fully functional on:

Two SCSI systems of the MCP2 XRC05 latest whatever level.

a.) One of which has the PlexWriter SCSI burner.
b.) One of which has an LG IDE burner.

One completely IDE system with MCP2 XRC05 latest whatever level.

a.) Which has an IDE hard drive Master on the first IDE controller.
b.) Which has a Master CD-ROM and a Slave LG DVD burner on the
second IDE controller

One completely IDE system with MCP2 XR05 latest whatever level.

a.) This is the ASUS Coppermine Intel CPU
b.) Which has no burner at all yet.
c.) But at least will now run Dani's drivers and everything so far
with no memory creep issues!

And EXACTLY the same system on Warp 4 level, with either Dani's
latest tools, or her last 1.7.10 DaniS506.ADD release, or IBM's
latest IBM1S506.ADD, will *NOT* run except like molasses. That
with FP17, TCP 4.3 latest, everything, that works fine on other
than this ASUS Coppermine box!

So it is *NOT* Dani's code which is the problem here, as far as I can see at all. It is simple that no form of Warp 4 can run properly on this motherboard and CPU without the memory access problem. PCI100 or PCI133 memory; whatever!

I still have research to do on the IBM ThinkPad issues yet for all this, which include USB backup hard drive capability and actual boot and re-clone, plus pure boot from the external USB system which works for me, prior to the latest Dani's tools. But thats gotta be done carefully as system changes if something goes wrong on the boot run have to be coordinated from floppy diskette boot runs for me. And I do not want to be a lonely puppy in the dark on this one.

Thanks Peter!

--


--> Sleep well; OS2's still awake! ;)

Mike Luther
.

Relevant Pages

  • Re: Strange PEER error with Danis 506 1.81 generic question
    ... First, a really large number of the systems around me are still SCSI systems which use not only SCSI hard drives, but also SCSI DAT drive tape backup systems, additional SCSI hard drives which can be plugged in, but are not normally there, in mobile drive trays; as well as, in some cases actual SCSI CD-RW units. ... The only use there for the IDE interface is another CD or DVD device for read purposes. ...
    (comp.os.os2.bugs)
  • Re: Strange PEER error with Danis 506 1.81 generic question
    ... First, a really large number of the systems around me are still SCSI systems which use not only SCSI hard drives, but also SCSI DAT drive tape backup systems, additional SCSI hard drives which can be plugged in, but are not normally there, in mobile drive trays; as well as, in some cases actual SCSI CD-RW units. ... The only use there for the IDE interface is another CD or DVD device for read purposes. ... But that flies in the face about data I/O speeds, as well as the requirement, as I understand things, that the CD device has to be a MASTER to be used for boot purposes if you are using, for example, the IBM install CD-ROM disks, or a modified version of this technique for custom boot CD-ROM operations. ...
    (comp.os.os2.bugs)
  • Re: Difference between IDE and SCSI ??
    ... IDE was one of the early standards. ... There was a half step between IDE and SCSI called ESDI. ... where the drives plug directly into the frame. ...
    (Fedora)
  • Re: Installed New IDE HD, Now Fedora Wont Boot from SCSI HD
    ... Now the computer wants to boot from the IDE drive, but the SCSI HD ... Linux started with SCSI hard drives so I think is should still work, ...
    (comp.os.linux.misc)
  • New hard disk appears in BIOS and Device Manager, not under My Computer
    ... IDE and SCSI hard drives don't like to play together. ... >On the IDE controller I also have a cd drive player. ...
    (microsoft.public.windowsxp.general)