Silicon Image 680 ATA 133 Data Corruption ???
 

Hi ALL !

I installed Clarkconnect / Redhat 9.0 Core on my old PC, P3 900, Intel 440 BX with a Silicon Image 680 (Sil 680 ) Ata 133 raid card and a 20 Gb IBM hard disc. It did run without problems.

Now i bought a Samsung 160 Gb Hard disk and, cloned the 20 GB IBM to the Samsung and connected the Samsung to the Sil680 to use full capacity of it.

AFter I made partitions of the remaining 140 GB with Partition Magic 8 the linux was unable to boot because of kernel errors due reading problems on the root partition (that i did not change anyway) and i fixed it by cloning this partition again from the IBM source disk.

The next issue was that a fresh formatted (extended) ext3 partition was corrupt (during automaount) after copying some mp3s through samba server. I had to reformat it with ext3 and copy the mp3s again. Then it works!

Now my questions:

1. Is it normal that with "fsck -nf" all of my disks are non-contiguous or " WARNING: Filesystem still has errors" , although the pc was never switched off suddenly and the partitions were particulary empty and some even a few hours old?

2. How (tool, programm) can I get rid off this hard disk erros?

3. Does the disk corruption depend on the use of the silicon image ata 133 raid controller ( I could use instead of it a Highpoint 370)?

I do NOT want one of my partitions (with stored mp3s or personal files) one day become unreadable!

4. Can I use the full capacity of the 160 GB hard disk with Linux or another OS when the harddisk is connected to the onboard controller, though the Bios does only recognize 128 GB (128 GB 32 Bit limitation) of the hd?