0

I am working with Ubuntu server 12.04 and earlier this week we detected a bad disk in /dev/sdl. I was not able to query its mount point and it also had failed SMART tests.

Today I brought a brand new hard disk to replace the failed one (same brand and model)

However after installing the drive, I run:

  $ dmesg -t | tail

to see if the drive has been detected. Unfortunately I am not able to detect the drive.

Here is the output of lshw | grep -A 5 disk.

  disk:10 UNCLAIMED
  description: SCSI Disk
  physical id: 0.a.0
  bus info: scsi@6:0.10.0

How can this appear regardless of whether I have a disk in there or not?

My process for replacing failed disks is this:

  • Physically locate bad disk
  • Unmount and remove bad disk
  • Insert new disk, run dmesg | tail to show disk is present
  • Format disk
  • Mount disk

For all the other disks, dmesg showed when they were inserted but not this time. Any thoughts?

Giovanni Tirloni
  • 5,693
  • 3
  • 24
  • 49
Brian Le
  • 1
  • 1
  • There are a [large number of ways](http://serverfault.com/questions/190685/whats-the-best-way-to-get-info-about-currently-unmounted-drives/190700#190700) to get information about drives, scanning the output of dmesg is a relatively poor method. – Zoredache Sep 12 '14 at 18:50

1 Answers1

0

Due to the age of this question, it may be safe to assume the system's been rebooted. However, there's another way to trigger a re-scan of the either the entire SCSI bus or just the device itself.

  • Re-Scan the entire SCSI bus

    echo "- - -" > /sys/class/scsi_host/host0/scan

  • Re-Scan just the device itself

    echo 1 > /sys/class/scsi_device/0:0:0:0/rescan

Note: You will have to confirm if host0 or 0:0:0:0 are correct for that host/drive.

Signal15
  • 943
  • 7
  • 27