---
title: Scrappy notes about RAID & mdadm
date: 2020-08-27
---
{% extends 'templates/base.html' %}
{% block body %}
{{ title }}
{% markdown %}
## Recover from a disk failure
Below is "what worked last time", but there is more comprehensive documentation [elsewhere](https://raid.wiki.kernel.org/index.php/RAID_Recovery).
Disconnect the existing RAID array:
```
$ mdadm --stop /dev/md127
```
Create a new array out of the working disks:
```
$ mdadm \
--verbose \
--assemble \
--force \
/dev/md0 \
/dev/sdc1 /dev/sdd1
```
Try mounting it, check everything's OK:
```
$ mount /dev/md0 /mnt/md0
```
Add the new disk:
```
$ mdadm --manage /dev/md0 --add /dev/sdb1
```
Watch it rebuild, and hope:
```
$ cat /proc/mdstat
```
## Misc
### Identify a physical disk
```
$ hdparm -I /dev/sda | grep Serial
```
{% endmarkdown %}
{% endblock %}