I have the following:
root@rharris:~# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid5]
read_ahead 1024 sectors
md0 : active raid1 lvmb[0] lvm-vg2/lvm2[1](F)
35413952 blocks [2/1] [U_]
md1 : active raid1 scsi/host0/bus0/target1/lun0/part1[0]
scsi/host0/bus0/target0/lun0/part1[1]
208768 blocks [2/2] [UU]
unused devices: <none>
root@rharris:~# uname -a
Linux rharris-build2.acs.internap.com 2.4.21-ac3 #15 SMP Wed Jul 2
12:13:51 EDT 2003 i686 unknown
sda1 and sdb1 are 200Meg partitions for /
lvm-vg1/lvm1 is a stripe of sda2+sdc2
lvm-vg2/lvm2 is a stripe of sdb2+sdd2
I physically hot pulled sdb out of the machine expecting
lvm-vg2/lvm2(md0) and the second mirror of md1 to fail. The md0 mirror
broke within seconds while the md1 mirror stayed intact for over 30 mins
even with me writing 50Meg and 100Meg files to /. It didn't mark sdb1
failed until I did a "pvscan".
This just strikes me as a bit odd that it wouldn't fail much sooner as
well. It would make sense if the root partition was 100% idle however I
did extensive reads and writes to / directly.
Robert
:wq!
---------------------------------------------------------------------------
Robert L. Harris | GPG Key ID: E344DA3B
@ x-hkp://pgp.mit.edu
DISCLAIMER:
These are MY OPINIONS ALONE. I speak for no-one else.
Diagnosis: witzelsucht
IPv6 = robert@ipv6.rdlg.net http://ipv6.rdlg.net
IPv4 = robert@mail.rdlg.net http://www.rdlg.net
Attachment:
pgp00051.pgp
Description: PGP signature
