On Saturday October 12, flux@modeemi.fi wrote: ... > > After moving the data from the 'failed disk' to the raid, I tried to > raidhotadd the failed device, however the operation failed as the disk was > too small. A little bit of a surprise to me, because the disks were the same > model, but the geometry was different, hence there was a few blocks > difference.. > > While trying to do that, I accidently added the whole disk to the raid. Of > course I tried to raidhotremove it, but that couldn't be done anymore, I > assumed due to the fact that it was already rebuilding the disk. Yes. raidhotremove will only remove an inactive spare or a failed device. Once rebuilding has started, you have to cause the device to fail before it can be removed: raidsetfaulty /dev/md0 /dev/hdX raidhotremove /dev/md0 /dev/hdX > > But I couldn't raithotremove it 800 minutes later either, which led me to > read the kernel logs, where I discovered the bug-statement. > > Here follows a ~50-line exerpt from the log. The complete log is available > at http://www.modeemi.fi/~flux/kern.log . The time (19:31) fits nicely to > the time I started the raid and then tried to remove it. This can be seen in > the full log. > > md: bug in file md.c, line 2351 That MD_BUG shouldn't really be there. It is just saying that you tried to remove a device that was busy. That should cause an error, but not a MD_BUG... however MD_BUG is not as catastrophic as BUG, all you get is annoying (and confusing) error messages. NeilBrown - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
