vamythguy wrote: > On 1/1/07, Curtis Doty <Curtis@xxxxxxxxxxxx> wrote: >> >> vamythguy wrote: >> > I have a RAID-5 array with four SATA disks in it. Apparently, two are >> > going >> > bad, so I want to replace them. But, I'm not sure which physical >> > drives to >> > pull. Given the following: >> > >> > cat /proc/mdstat >> > Personalities : [raid6] [raid5] [raid4] >> > md0 : active raid5 sda[0] sdd[3] sdc[2] sdb[1] >> > 735351936 blocks level 5, 64k chunk, algorithm 2 [4/4] [UUUU] >> > >> > Any suggestions? >> >> Odd that you have the raid4 and raid6 modules loaded, but still >> unfamiliar with /proc/mdstat... >> >> Anyways, you are showing an array with four working disks; none failed. >> If you had a failed drive, it might look like [U_UU] which would mean >> the second disk had failed. >> > > I know how /proc/mdstat works, so I know the drives haven't failed *yet*. > > However my system log is reporting: > > Dec 31 17:04:38 tibeaux smartd[2384]: Device: /dev/sda, 3 Offline > uncorrectable sectors > Dec 31 17:04:39 tibeaux smartd[2384]: Device: /dev/sdd, 48 Currently > unreadable (pending) sectors > > over and over again. > > And another list member (Pedro Fernandes Macedo < > webmaster@xxxxxxxxxxxxxxxxxxx>) says this means the two disks are in > danger > of failing. So, unless I can safely prevent against that, I want to > replace the disks before they do. I would just add a hot spare and let > mdadm do it, but there is no room in the case. So, back to the original > question... > Those log entries are coming from smartd, not mdadm. You can always fail one disk out of the array with mdadm /dev/md0 -f /dev/sde and then hot add another one back later. However, smartd is throwing errors on two disks and multi-disk failure with raid5 means catastrophe. Since mdadm hasn't already failed one automatically, I'd first look more closely at smartd and what preemptive information in can report. Is your original question simply how to identify /dev/sda and /dev/sdd? ../C
