Hi,
I've been playing around with software raid (raid1) using ide drives
(hda, hdc) with kernel 2.4.18 on Trustix 1.5. To simulate a drive
failure, one of the drives (hdc) was removed to create a degraded mirror
situation. I then proceeded to use the remaining drive for a while. I
then plugged the "failed" drive back in and rebooted. The kernel chose
to boot from the failed drive over the "good" drive thus all new changes
that were on the good drive were lost during resyncing of the drive
pair. In this case the good drive was marked dirty. Make sense? Any
suggestions on how to ensure the "bad" drive (ie. older data) gets
marked as bad so the mirroring process overwrites it instead? Do I have
to reformat/erase the pulled-out drive to preserve the most current
data?
Prior to raidhotadding the bad drive into the raid:
#cat /proc/mdstat
Personalities : [raid0] [raid1] [raid5]
read_ahead 1024 sectors
md3 : active raid1 hdc1[0]
10432 blocks [2/1] [U_]
md2 : active raid1 hdc2[0]
262464 blocks [2/1] [U_]
md1 : active raid1 hdc5[0]
204992 blocks [2/1] [U_]
md0 : active raid1 hdc6[0]
777024 blocks [2/1] [U_]
unused devices: <none>
------------------------------
Here's some mdadm (only used for diagnostic purposes) output after
replacing the "bad" drive:
#mdadm --examine /dev/hdc1
/dev/hdc1:
Magic : a92b4efc
Version : 00.90.00
UUID : 2a727c39:32f6a941:6ed0f7f1:d23d1eda
Creation Time : Tue Mar 5 10:04:44 2002
Raid Level : raid1
Device Size : 10432 (10.18 MiB 10.68 MB)
Raid Disks : 2
Total Disks : 1
Preferred Minor : 3
Update Time : Sat Mar 30 02:39:56 2002
State : dirty, no-errors
Active Drives : 1
Working Drives : 1
Failed Drives : 0
Spare Drives : 0
Checksum : c0cdd422 - correct
Events : 0.82
Number Major Minor RaidDisk State
this 0 22 1 0 active sync /dev/hdc1
0 0 22 1 0 active sync /dev/hdc1
1 1 0 0 1 faulty removed
-----------------
#mdadm --examine --scan /dev/hdc1
/dev/hdc1:
Magic : a92b4efc
Version : 00.90.00
UUID : 2a727c39:32f6a941:6ed0f7f1:d23d1eda
Creation Time : Tue Mar 5 10:04:44 2002
Raid Level : raid1
Device Size : 10432 (10.18 MiB 10.68 MB)
Raid Disks : 2
Total Disks : 1
Preferred Minor : 3
Update Time : Sat Mar 30 02:39:56 2002
State : dirty, no-errors
Active Drives : 1
Working Drives : 1
Failed Drives : 0
Spare Drives : 0
Checksum : c0cdd422 - correct
Events : 0.82
Number Major Minor RaidDisk State
this 0 22 1 0 active sync /dev/hdc1
0 0 22 1 0 active sync /dev/hdc1
1 1 0 0 1 faulty removed
Thanks,
Stephen
-
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
