Ji I am having tons of trouble with this damn NT4 server, the files are getting corrupted only after 2 days, I think I must have done something wrong (this is my first NT4 so please excuse my bad understanding of the product). I did a search and found the following quotes, I cant seem to work out what it meant by first create a 'SYSTEM' partition and then the 'BOOT' partition, I just read through the WinNT4 pro reference book and cant find any references to it either. My current set up is 4GB SCSI on NTFS (single partition on a 9GB SCSI drive) and data are held on a 18GB SCSI drive (ID-2). You think I did something wrong ? as I have been scracting my head for weeks now, I cant talk them into Win2000 server, but I am guessing it has something to do with the way I set up the drives and partitions thats causing this file corruption. If anyone can help, I would be most grateful. ============================================ [ QUOTE ] Before starting the installation insert the disk into an existing NT installation and partition/format the disk using Disk Administrator and then insert the disk into the machine to be installed Partition the disk into smaller partitions, if you had a 5GB disk you could have a 1GB system partition, and a 4GB boot partition. The system partition is the partition NT's core startup files are located, boot.ini, ntldr and ntdetect.com (ntbootdd.sys if SCSI), and will normally be the active partition. The boot partition is the partition that NT stores the rest of its files, i.e. the %systemroot% directory Create a 4GB partition at installation time, and then extend the NTFS partition after installation has completed - Start Disk Administrator (Start - Programs - Administrative Tools - Disk Administrator) - Select the NTFS partition and holding down the Ctrl key select the unpartitioned space of the rest of the disk - From the Partition menu, select Extend Volume Set Note - You cannot extend a NTFS partition if it is the boot or system partition (as the boot/system partition cannot be part of a volume set)

It sounds to me that you have some nasty low-level hardware problem probably on the main data bus of the motherboard. Data would appear the be getting corrupted before writing to/after reading from the disc system. When data is written to the disc the hardware generates checksum characters and when it reads data the checksum is check to make sure the data has been read ok. If the data is getting corrupted prior to writing to disk then the corrupted data will be getting written away to disk etc. If this system has a diagnostic boot partition I would booty into and start running some extensive diagnostic tests to see if the tests pick anything up. But I would not trust it with 'live' data until you have managed to identify the cause of the problem.

Thanks Maurice I checked both drives before installation, in fact, I ran a SCSI format using the built-in function of the 29160N BIOS and found no errors. Only thing I can think of is when I first install NT4, the message of NT4 and >1024 cylinders, which I ignored. I think you are correct on this, as I found this morning, the Tapeware software had died after yet another failed backup, on trying to start up the program again, it first came up with the message that the system is low on virtual memory (Drive C still had 3GB of free space), then it hanged the system, after restarting, Tapeware would not start up. I am now even more baffled then last week...

If I recall you have run chkdsk and it has never found errors. It is only after a bit writing files that data becomes corrupt and that is internal to the files themselves. I think you might be looking the wrong subsystem. Who put this box together? In otherwords is the ram correct for the specs of the motherboard? If you have mismatched or slightly compatable ram it can manifest itself in data corruption. I also don't recall what drives types these are and how you are terminated. Improper termination on the scsi bus can also lead to this kind of erratic corruption. Let me give you an example. Adaptec 3200s raid card. Seagate scsi2 drives. I have a active terminator at the end of the scsi cable for the 7 drives [two cables two busses off the card each with a terminator. three drives on one chain, 4 drives on the other]. I go and make a raid 0+1 array. It works for a day or so but I have random drives drop out so the system uses the hot spare and rebuilds on the fly. What the heck is going on? So I do some research and find that seagate drives come preconfigured with termination enabled !!! Add to this that the jumper is on the underside of the drive !!! Well you can't have drives terminated and a active terminator. According to scsi rules you can only have the card and the end of the bus terminated. So I remove the jumpers from all of the drives and walla problem solved. So I ask where/how/who put the hardware together

Thanks for the suggestion of RAM and Terminator. This box came 2 years ago, pre-configured by NEC, it had been working perfectly until 2 months ago, user reported a virus, then a maintenence company came to clean the workstation, then this lead to a server re-build, plus install NAV on , the whole process took 58 hrs. After that, they still had 3 WK infected with the W32.pinfi virus, so the box came to us, and for 3 weeks, I have been trying and trying, but now after your suggestions, it could just be that, they added extra memory after the rebuild, and also messed around with the SCSI when adding a 5400rpm IDE to hold the daily use data. I were suspecting the problems came from the server hardware/software itself, but just werent able to pinpoint the cause of it, but now I have something to work with, I will try to hold the data together until Friday when I can take the box offsite and give it another bash. Cant thank you enough for helping, most grateful.