2x NT 4.0 workstation on same machine; BOOT.INI syntax?? HAL ?? ARC??

[artwork]

her's my problem: I have to IDE drives, one on prim. and one on sec. IDE controller. (both with a primary partition and several logical part.) <br>

I want to set up WIN NT twice, once on every disk, so that in case of crash all I have to do is edit the boot file, and I'm ready to continue (after removing one of the HD's. <br>

What I did was: <br>

-disconnect cable on Sec. IDE controller, istall WIN NT 4 on Prim. part. on first IDE controller. <br>

-repeated this for the sec. IDE controller (disconnect cable and install NT) <br>

- in Partition Magic, I hide the primary partition of the second HD (IDE2), change letter for primary part. of first IDE controller into C:, unhide primary part of IDE2 controller and change letter into F:, <br>

so now I have: <br>

c:\ full NT 4. install. <br>

f:\ full NT 4. install. <br>

This seemed better than installing NT in the normal way twice, since I found out that NT is NOT installing all the files. E.g. NTOSKRNL.Exe is not installed, no booti.ini, no program files folder etc., so I'm afraid I can't just rip out one of the HD's and continue. <br>

<br>

Problem: <br>

What is the correct syntax for the boot.ini? Have tried everything (multi (0), (1), (2) and also the same for the other options (Rdisk etc.), but nothing works, either WIN can't find NTOSKRNL.exe or it states

 

[MasterRacker]

The way I understand the boot process, multiple installs will still leave only one boot manager on your primarty disk. If that fails, you have no boot manager to redirect you to a second disk.<br>
<br>
I don't quite understand why you're going through all this. Are you storing data files on the mahcine as well, our only putting data on a server? If you're in a required high-availability situation and your data must be stored locally I would do the following:<br>
<br>
1. Install NT and all apps onto the primary drive. <br>
2. Insert a second disk and clone the primary using Ghost or something similar.<br>
3. Remove and store the clone.<br>
4. Insert and format the data disk.<br>
<br>
In the event of a failure, spend 5 min. with a screw driver to replace the primary and you are running again. <p> Jeff<br><a href=mailto: masterracker@hotmail.com> masterracker@hotmail.com</a><br><a href= > </a><br>

 

[BruceP]

Here is how it should look. The first line should be the primary hard drive (drive c) and the next hard drive should be the backup (drive f).<br>
<br>
Multi(0) should always be zero unless you are using multiple scsi adapters. disk(0) should always be zero unless you are using a scsi adapter. Rdisk(0) is the drive numbers starting at zero. Partition(1) is the primary partition of that particular hard drive.<br>
<br>
multi(0)disk(0)rdisk(0)partition(1)\WINNT<br>
multi(0)disk(0)rdisk(1)partition(1)\WINNT<br>
<br>
Instead of doing all of this, why not configure disk mirroring on the machine. This way if a drive fails you can break the mirror, replace it with a new unit, and re-establish a mirror. <br>

 

[MasterRacker]

Good one Bruce. Mirroring is another good option. <br>
<br>
A good question here is this? Why bother? Redundancy and fault tolerance are an absolute necessity on a server, however, theoretically all data should be on a server and the hard disks in a workstation are there only to hold an OS and applications. This way you get centralized backup, file sharing, and a ton of other benefits. Even if you need more speed, you pull down a file work on it locally and put it back on the server.<br>
<br>
Granting all this, how often does a hard disk actually fail these days? How much time and extra hardware expense do you expend building redundancy into workstations versus using cloning or a couple of spare hard disks to rebuild the rare catastrophic failure? <p> Jeff<br><a href=mailto: masterracker@hotmail.com> masterracker@hotmail.com</a><br><a href= > </a><br>