This is a discussion on Disk Druid/Drive Image within the Linux Operating System forums, part of the Unix Operating Systems category; --> Have 2 x 60Gb drives. Drive1 has one NTFS partition on which I have XP. In addition there are ...
| |||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| Have 2 x 60Gb drives. Drive1 has one NTFS partition on which I have XP. In addition there are two other FAT32 partitions. Drive2 contains two ext3 partititions and 1 swap (totalling 7GB), followed by three FAT32 partitions. Prior to installing RH9.0 I always used Drive Image to create backup images of my main C drive, ie the NTFS partition. However after installing RH9.0, Drive Image flagged up #116 errors on both drives, on all partitions starting after cyl 1023. It stated that the LBA address did not match the CHS address and DI could 'fix' it, the LBA being deemed correct in all cases.. I allowed DI to do the 'fix' and all now seems OK, both for Linux and XP. Oddly, the CHS address given by DI as the start address of all affected partitions was 16450559. All partitions clearly could not start at the same address, and I suspect that 16450559 is equivalent to 1024,255,63, in other words the CHS limit of the bios. Anyway, Drive Image, like Partition Magic, insists upon a match between LBA and CHS in each partition table before it can continue. Thus I assume that the 'fix' was to enter the correct CHS address in each partition table, even though higher values could not be used by the bios. But as I understand it, modern systems use the EDD standard (Enhanced Disk Drive) where only the LBA address are used; the CHS address being essentially redundant. Thus would I be correct in assuming that the 'fix' did no more than satisfy Drive Image's need for a match, and should have no effect upon the operation of XP or Linux? Certainly to date, I have not seen any adverse effect on either operating system. Secondly, these 'errors' were introduced during the installation of RH9.0, my first Linux installation. Prior to the install I created a 7GB unallocated space at the start of the second drive, then during the installation I allowed RH9 to create the 3 partitions required automatically. This suggests it all has something to do with Disk Druid which Redhat uses. I have created/resized partitions on numerous occasions using MS's Fdisk or Partition Magic, and have had no problems at all with either DI or PM. Thus Disk Druid must do something differently to those other partition managers. I also find it strange that both drives should be affected in the same way. I deliberately chose to install RH on drive2 and also boot via floppy in order that drive1, ie my XP drive, remained untouched. Clearly Disk Druid does cross drive boundaries, although being told, in my case, to create the necessary partitions in the unallocated space on hdb. This is a somewhat worrying characteristic. This same #116 problem has been discussed several times. It also apparently occurs after the install of Suse, which I assume also uses Disk Druid during its install. Many people simply say do not 'fix' in case something gets screwed up and you risk loss of access to the drives, but if one took that advice DI or PM could not be used, which is a pity as both are very useful applications. Can anyone throw some light on this? |