it means that your system will not crash. no data loss would happen The problem i had was that the system kept on bouncing io to the defected disk and since both the root file system and the swap were on raid1 the system hanged after a short while. I added a line that removed the defected ata port ( the disk ) from the disks array in the kernel and by doing that i had prevented this io bouncing. On Sun, 2005-07-03 at 18:43, Frank Wittig wrote: > with current version 2.6 kernels there are no problems known to me not > to have swap and system raid partitions built out of partitions of the > same physical disks. > i didn't catch what you mean by "the kernel that removes the bad disk > from the operating system". if it means that it will trigger a swapoff > for the partition on the failed disk it is near to useless since at that > time already a irreversible data loss has happened which causes the > system to crash. > the only possibility to prevent a system crash during failure of a > harddisk is swapping to raid partitions. > > raz ben jehuda wrote: > > i had a "system hanging" when both swap and root file system were on the > > raid. for that i added line in the kernel that removes the bad disk from > > the operating system. > > On Thu, 2005-06-30 at 22:53, Frank Wittig wrote: > > > >>Eric Pretorious wrote: > >> > >>>On Thursday 30 June 2005 12:03 pm, Laurent CARON wrote: > >>> > >>> > >>>>is the swap partition on raid? > >>> > >>> > >>>No. swap both swap partitions are formatted as standard swap partitions and mounted via /etc/fstab. > >>> > >> > >>That's no good idea. If a disk fails which contains a swap partition to > >>which data has been swapped the system will crash. > >>Therefor it is important to use mirrored partitions for swap. -- Raz Long Live The Penguin - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
