Hi,
I have a system with two IDE controllers running RAID1. As a test I powered down, removed one drive (hdc), and powered back up. System came up fine, so powered down installed a new drive (hdc) And powered back up. /proc/mdstat indicatd RAID1 active with hda only. I thought it would Auto add the new hdc drive... Also when I removed the new drive and added The original hdc, the swap partitions were active hda and hdc but only hda on the other partitions. I has to add the other hdc partitions with mdadm -a.
My mdadm.conf looks like;
# mdadm.conf written out by anaconda DEVICE partitions MAILADDR root ARRAY /dev/md1 super-minor=1 ARRAY /dev/md0 super-minor=0 ARRAY /dev/md3 super-minor=3 ARRAY /dev/md2 super-minor=2
Shouldn't there be more information for mdadm to work with? How do you replace a failed drive and have it auto-configured?
TIA Gerald
You need to "reinsert" it manually. It's very well explained in this Howto: http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html See the section 6.3 to know how to add the new drive to the array.
I tested this 3 days ago :)
----- Original Message ----- From: "Gerald Waugh" gwaugh@frontstreetnetworks.com To: "CentOS mailing list" centos@centos.org Sent: Saturday, May 21, 2005 5:09 PM Subject: [CentOS] Software RAID CentOS4
Hi,
I have a system with two IDE controllers running RAID1. As a test I powered down, removed one drive (hdc), and powered back up. System came up fine, so powered down installed a new drive (hdc) And powered back up. /proc/mdstat indicatd RAID1 active with hda only. I thought it would Auto add the new hdc drive... Also when I removed the new drive and added The original hdc, the swap partitions were active hda and hdc but only hda on the other partitions. I has to add the other hdc partitions with mdadm -a.
My mdadm.conf looks like;
# mdadm.conf written out by anaconda DEVICE partitions MAILADDR root ARRAY /dev/md1 super-minor=1 ARRAY /dev/md0 super-minor=0 ARRAY /dev/md3 super-minor=3 ARRAY /dev/md2 super-minor=2
Shouldn't there be more information for mdadm to work with? How do you replace a failed drive and have it auto-configured?
TIA Gerald
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
You need to "reinsert" it manually. It's very well explained in this Howto: http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html See the section 6.3 to know how to add the new drive to the array.
I tested this 3 days ago :)
There has to be a better way than adding manually. We ran RAID1 on 2.2 kernels for years, and anybody, could replace a drive And it would be added automagically...
Maybe adding "auto" to mdadm.conf and better describing the devices.
Gerald
On 5/21/05, Gerald Waugh gwaugh@frontstreetnetworks.com wrote:
You need to "reinsert" it manually. It's very well explained in this Howto: http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html See the section 6.3 to know how to add the new drive to the array.
I tested this 3 days ago :)
There has to be a better way than adding manually. We ran RAID1 on 2.2 kernels for years, and anybody, could replace a drive And it would be added automagically...
Maybe adding "auto" to mdadm.conf and better describing the devices.
How is the system supposed to know that a once-broken drive is now not broken? You can have other failures beside disconnecting. Or which, if any, of your running partitions should overwrite the data on a freshly added disk that might already contain needed data? You can pre-configure spare devices if you want them to automatically replace failed ones but otherwise the system shouldn't do potentially destructive changes by itself.
-
Gerald Waugh -
Les Mikesell -
Steven Moix - Axianet.ch