Begging for some insights from someone who may have had a similar problem:
I spent most of yesterday with the guy who built my PII350 machine trying to add a second hard drive. The original one was a Maxtor model 90680D4 (6.8GB). The one we were adding is a Western Digital WDC64AA (6.4GB).
Many strange things happened along the way, but at the end of the day I had one drive working, the WDC, and the Maxtor sitting outside. It seemed like nothing we could do would allow the drives to work together.
I have a 1BIT BH6 motherboard with two IDE connectors, the hard disk as the Primary Master, and an ATAPI CD ROM drive as the Secondary Slave.
The first thing we did was to install the WDC as the Primary Master, and the Maxtor as the Primary Slave and then try a diskcopy from the old drive to the new, believing the WDC to be a bit faster, and because I was a bit nervous about the Maxtor in the first place. (Last week I was pushed into scandisk at startup, something I had been lax about running when I should. After a lengthy session, I had over 700KB of bad area marked on the disk. The next day I got the first boot failure I have ever seen. It did boot after a power cycle, and seemed fine after that. Also, a defrag finished without a hitch.) The drive is partitioned into a 4+ GB and a 2GB, which usually show up as C: and D:
The diskcopy did not complete. The screen was showing that things stopped during the D: partition copy, probably at the very beginning. However, we were able to boot up from the Maxtor at that point. (We made it the master again figuring the WDC was not right) Then we looked at the contents of the drives and found that D: was identical to C:. I don't recall even looking at E:, but that was a bit of a panic thinking that everything I had on D: was forever lost, replaced by a second copy of what had been on C:
Because of boot problems with 2 drives in place, we looked at the WDC installed alone. It appeared to be a full copy of the old C: partition, and it had an empty D: partition. Only the sizes of the partitions were showing as the sizes of the original drive, which is physically impossible. After fooling around with the partitions and reformatting the D: partition, we had what looked to be a valid C: partition with all the transferred files, and an empty D: partition of 1.7GB, which looks right.
Next step was to install the old Maxtor in another machine in place of an existing second drive. (The guy has a nice little drawer setup for that, so you can change drives from outside the cabinet.) After getting the jumper on the drive set to what appeared to be the wrong setting, based on the words on the drive, the disk appeared to be perfectly normal showing as drives D: and E: in his machine, with all the original data intact. Since we already had the original C: (now showing as D:) copied, and because it would fit, he burned a CD ROM copy of the original D: partition (now showing as E:), which we used to transfer the files to the D: partition of the WDC. So now we had a good (we think) WDC drive, and a good Maxtor drive (we think) that except for a small size difference are mirror copies of one another.
Next step was to get the two drives working together, being aware that the jumper settings printed on the drives might not be right. With both drives in place things were very unreliable. It might boot up, it might not. If it did boot up at all, the drive designations were interlaced: C: was the first partition of the Primary Master, D: was the first partition of the Primary Slave E: was the second partition of the Master, etc. (This probably accounts for why after the original copy failed, it looked like C: had been duplicated on the second partition of the original drive. We assumed D: was the second sector, but in fact it was probably the first sector of the new drive.)
After a reasonably careful check of the new WDC, believing it was a valid duplicate of the old Maxtor, we did a complete reformat of the Maxtor as a single partition drive, as a bootable disk and loaded it with Windows98, assuming that perhaps there really was a problem with the Maxtor that had caused the original diskcopy to stall. Still there was no way to get drives to work reliably together. When we were able to get the system up, the drive designations were still interlaced. We took the Maxtor out and left it with the single WDC running with the C: and D: partitions. So far, so good.
We ran a scandisk on the WDC (the standard version) and at the end it was showing the same number of bad bytes on partition C: as had been on the old disk. The D: partition (the one that had been reformatted) was clean even though the original disk had 4096 bad bytes. This raises a question about how bad clusters get marked on the disk. It seems the WDC picked up a table of bad clusters from the original copy from the Maxtor, but only on the C: partition. The scandisk found no new errors, but reported all the bad space, so I assume it never checked the clusters that had been bad on the original disk. Is there something within the drive that marks bad clusters, or is that information only in a system file? If I reformat a disk, do I have to scandisk to exclude bad areas again? If I run a complete scandisk on the new WDC, will it check the entire disk and perhaps "recover" the clusters that are marked bad but are really OK?
I know this is long, but I tried to tell the whole story. Any help would be much appreciated.
Dan |
