How to detect new hard disk attached without rebooting in Linux

by

 

How to detect new hard disk

Here suppose you have added a new hard disk to your Linux OS running on any Virtual Environment which in my case is VMware workstation. Once added the new hard drive, generally the changes won't reflect unless you reboot the Guest OS.
But what if you are in no position to reboot the Guest Linux OS?
 

Solution

In the below path you can find a list of host symlinks pointing to the iscsi device configured on your Linux box

# ls -l /sys/class/scsi_host/
total 0
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host0 -> ../../devices/pci0000:00/0000:00:07.1/host0/scsi_host/host0
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host1 -> ../../devices/pci0000:00/0000:00:07.1/host1/scsi_host/host1
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host10 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host10/scsi_host/host10
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host2 -> ../../devices/pci0000:00/0000:00:10.0/host2/scsi_host/host2
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host3 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host3/scsi_host/host3
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host4 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host4/scsi_host/host4
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host5 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host5/scsi_host/host5
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host6 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host6/scsi_host/host6
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host7 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host7/scsi_host/host7
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host8 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host8/scsi_host/host8
lrwxrwxrwx 1 root root 0 Jun 14 05:08 host9 -> ../../devices/pci0000:00/0000:00:11.0/0000:02:05.0/host9/scsi_host/host9

But to detect a new hard drive attached you need to first get your host bus number used which you can get by using below command

# grep mpt /sys/class/scsi_host/host?/proc_name

You should get a output like below

/sys/class/scsi_host/host2/proc_name:mptspi

So as you see your host2 is the relevant fields where you need to reset the storage buffer values. Run the below command

# echo "- - -" > /sys/class/scsi_host/host2/scan

Here "- - -" defines the three values stored inside host*/scan i.e. channel number, SCSI target ID, and LUN values. We are simply replacing the values with wild cards so that it can detect new changes attached to the Linux box. This procedure will add LUNs, but not remove them.
Once done verify if you can see the new hard drive which in my case worked very fine as I see below

# fdisk -l
Disk /dev/sdb: 3221 MB, 3221225472 bytes
 255 heads, 63 sectors/track, 391 cylinders
 Units = cylinders of 16065 * 512 = 8225280 bytes
 Sector size (logical/physical): 512 bytes / 512 bytes
 I/O size (minimum/optimal): 512 bytes / 512 bytes
 Disk identifier: 0x00000000

 
Let me know your success and failures.

15 thoughts on “How to detect new hard disk attached without rebooting in Linux”

  1. Hello Prasad,
    Quick tip:
    A better way to identify new disks is using the rescan-scsi-bus.sh script.
    You can install it with by 'yum install sg3_utils'.
    Some options you'll want to use are: -c or -cr.
    By the way, your 'Configure Red Hat Cluster using VMware, Quorum Disk, GFS2, Openfiler' is really good. Congrats and keep the good tutorials.

    Reply

  12. I had a permission denied error until I did the following on the echo line:
    sudo bash –c ‘echo “- – -“ > /sys/class/scsi_host/host2/scan’
    Then all was right with the world. Thanks for the very useful article 🙂

    Reply

Leave a Comment