Linux How-To Monitor DD Progress Copying A Drive to a New One

Linux How-To Use DD to Copy A Drive to a New One

There are times when it's necessary to migrate all the data from one disk to another. These include

• Upgrading to a larger disk
• Pre-emptivey moving to a new drive following a SMART alert
• Making a copy before moving a server

etc

There are lots of software tools that can be used, both open-source/free and paid-for, for example CloneZilla

One of the simplest is simply to use the Linux command 'dd' from a recovery disk

dd is a command-line tool for copying disk content (filesystems) from one location to another, primarily used to make a 'file' of a specific size for speed/io testing, or to copy an entire disk image.

WARNING: dd is often called data-destryer for good reason if you're not 100% sure on what you are doing !

As simple and quick as DD is, as with most CLI tools, it operates 'silently' unless there are errors to report - and so there are disadvantages for use as a drive duplication too, like being able to monitor the progress of the disk copy - below are the 2 most common solutions ...

How to show progress during a dd copy

Simple Method - using iostat

If you started your machine with a SysRescCD Live-Linux distribution (the most widely used way to get a box 'up' to be able to DD the contents - then you have access to another command-line tool called 'iostat' <p.> To check how far your disk duplication has reached, simply open another terminal window (ALT-F2) and type:
$ iostat

This will report the KB_read and kB_wrtn (written) for your disks - divide the written by 1024 and you have the amount of MB completed so far, and can guestimate the remaining time based on the size of the disk (which you can obtain from an 'fdisk -l'

Advanced Method - sending a USR1 kill to dd

For a more 'interactive' progress report you can send a USR1signal to the dd process from another terminal window ...

1. Determine the process-id (pid) for the running dd command:
   $ pgrep -l '^dd$'
   This will output the pid and binary name e.g. 2188 dd
2. Send the USR1 signal to the pid:
   $ kill -USR1 2188
   When dd picks up the USR1 signal it will print the current statistics to STDERR on the original terminal and continue.
3. Switch to the terminal running dd and view the output:
   e.g.
   12698193+0 records in
   12698193+0 records out
   54735860224 bytes (55 GB) copied, 4529.87 seconds, 12.1 MB/s
4. Use the 'watch' command to execute the USR1 kill every minute:
   $ watch -n 60 kill -USR1 2188
5. Kill the watch when the copy has completed with CTRL-C

Add to Favourites    Print this Article

How do I add an E-mail Account to Webmin (in Virtualmin) ? (Views: 9071)

Set up SSH for use without a password? (Views: 2081)

What are the default ports to open for a firewall ? (Views: 2694)

How to sort directories (folders) by size in Linux (Views: 7962)

Can you fix MySQL on my Unmanaged VPS or Unmanaged Root Dedicated-Server? (Views: 3836)