On 2012-01-09 at 09:42:17, Pádraig Brady wrote: > Highly unlikely. > If so it would be a kernel bug, but it's almost certain that is not what > happened. > disk enumeration is often quite complex. > Did you reboot after the wipe? Perhaps the disks are now swapped because > of your wiping operation?
What was really strange and can't explain is that everything started failing on /dev/sdb after the wipe of /dev/sda completed successfully. For example, running "top" or cat'ing a file would fail with "Input/Output Error". Which is why I started to suspect that perhaps the dd operation has jumped to the other device. Then it stopped booting after I power-cycled it. Anyways, it sounds like it's unrelated (maybe a loose cable or something). I will investigate further once I get the drives (it was in a remote server which I can no longer ssh into). > p.s. There is no need to specify count if write to the whole device. > Also bs=1M might be faster. That's good to know. So I guess I should have done this: dd if=/dev/zero of=/dev/sda bs=1M What's the expected behavior if a dd command causes dd to attempt to write past the end of the drive (I'm thinking that maybe I got my math wrong)? It just stops with an error of some sort? Cheers, Francois -- Francois Marier identi.ca/fmarier http://fmarier.org twitter.com/fmarier
