Differences

This shows you the differences between two versions of the page.

--- linux:centos:software-raid1-setup [2020/04/13 15:29] – odefta
+++ linux:centos:software-raid1-setup [2023/07/26 16:04] – [Re-add a good hard disk (after cable disconnection for example)] odefta
@@ Line 3: / Line 3: @@
 ====== Intro ======
-RAID 1 = mirrored pair of two hard disks
+RAID 1 = mirrored pair of two hard disks \\
-If one disk fail, the raid will continue to be operational.
+If one disk fail, the raid will continue to be operational. \\
-Read performance will be increased for parallel read operations.
+Read performance will be increased for parallel read operations. \\
-Write performance will be likely the same (almost the same).
+Write performance will be likely the same (almost the same). \\
 ====== Prerequisites ======
@@ Line 19: / Line 19: @@
 ====== Setup ======
-Format first hard disk (/dev/sda) with xfs (supports also hard disks higher than 2TB)
+==== Format first hard disk (/dev/sda) with xfs (supports also hard disks higher than 2TB) ====
 <code>
@@ Line 28: / Line 28: @@
 -L = hdd label
-Format second hard disk (/dev/sdb) with xfs.
+==== Format second hard disk (/dev/sdb) with xfs ====
 <code>
@@ Line 34: / Line 34: @@
 </code>
-Create the raid array.
+==== Create the raid array ====
 <code>
@@ Line 40: / Line 40: @@
 </code>
-Format the raid array as xfs.
+==== Format the raid array as xfs ====
 <code>
@@ Line 46: / Line 46: @@
 </code>
-Mount the raid array (in /data1 folder)
+==== Mount the raid array (in /data1 folder) ====
 <code>
@@ Line 56: / Line 56: @@
 </code>
-Automatically mount the raid volume at startup. \\
+==== Automatically mount the raid volume at startup ====
 Edit **/etc/fstab** and add:
 <code>
@@ Line 62: / Line 63: @@
 </code>
-Load automatically the raid 1 configuration at startup.
+==== Load automatically the raid 1 configuration at startup  ====
 <code>
 mdadm --detail --scan --verbose >> /etc/mdadm.conf
 </code>
-Check the raid configuration.
+==== Check the raid configuration ====
 <code>
 mdadm -D /dev/md0
@@ Line 105: / Line 108: @@
 </code>
+==== Replace a failed hard disk in the raid configuration ====
+Suppose the /dev/sdb hard disk failed. \\
+To replace it with a good hdd, set it as failed and remove it from the array:
+<code>
+mdadm /dev/md0 --fail /dev/sdb --remove /dev/sdb
+</code>
+Add the new device to the array:
+<code>
+mdadm --add /dev/md0 /dev/sdb
+</code>
+==== Re-add a good hard disk (after cable disconnection for example) ====
+After reinserting the hard disk, it may be mounted now on a different device letter.\\Ckeck it by running:
+<code>
+fdisk -l
+</code>
+Output:
+<code>
+...
+Disk /dev/sdd: 4000.8 GB, 4000787029504 bytes, 7814037167 sectors
+Units = sectors of 1 * 512 = 512 bytes
+Sector size (logical/physical): 512 bytes / 4096 bytes
+I/O size (minimum/optimal): 4096 bytes / 4096 bytes
+</code>
+After identifying it, re-add it to the raid array with:
+<code>
+mdadm --manage /dev/md0 --add /dev/sdd
+</code>