How to check dd progress in Linux

A raw image write can look frozen when dd has no progress output, especially during a large USB write, disk clone, or recovery copy. Checking transfer progress while the copy runs shows whether bytes are still moving before interrupting a job that may only be slow.

GNU dd from coreutils supports status=progress for new transfers. The option prints periodic byte counts and a final summary to standard error, so it works with ordinary file copies and block-device writes without adding another monitor command. Short test copies may finish before a visible intermediate update, but the final summary still confirms the byte count and speed.

For a dd process that is already running without status=progress, Linux systems with GNU dd can print a one-time report by sending USR1 to the selected process. The report appears in the terminal where dd is running. Progress checks do not make dd safer, so confirm both if= and of= before copying to disks, partitions, or installer media.

1. Run dd with status=progress for a new transfer.

   $ dd if=/dev/zero of=/tmp/dd-demo.img bs=1M count=64 status=progress conv=fsync
   64+0 records in
   64+0 records out
   67108864 bytes (67 MB, 64 MiB) copied, 0.188307 s, 357 MB/s

   Replace the sample if= and of= paths with the real source and destination. Use sudo when the source or destination is a block device such as /dev/sdb.

2. Confirm that the output file has the expected size.

   $ ls -lh /tmp/dd-demo.img
   -rw-r--r-- 1 user user 64M Jun 13 20:20 /tmp/dd-demo.img

   The final dd summary and the file size should agree. For a real disk or partition target, verify the written device with the guide that matches that workflow rather than relying only on the byte count.

3. Start a temporary background dd copy only when no real copy is already running.

   $ dd if=/dev/zero of=/tmp/dd-fallback.img bs=1M count=1024 oflag=sync &
   [1] 31244

   Skip this step when monitoring an existing long-running dd process. The oflag=sync sample slows the temporary file write enough to test the signal path on many systems.

4. Identify the running dd process from another terminal before sending a signal.

   $ pgrep -a '^dd$'
   31244 dd if=/dev/zero of=/tmp/dd-fallback.img bs=1M count=1024 oflag=sync

   Select the process ID for the exact copy you intend to monitor. Use sudo pgrep -a '^dd$' when the process belongs to another user.

5. Send USR1 to the selected dd process.

   $ kill -USR1 31244

   Use sudo kill -USR1 31244 when the process is owned by root. The progress report appears in the terminal where dd is running.

6. Read the progress report in the original dd terminal.

   287+0 records in
   287+0 records out
   300941312 bytes (301 MB, 287 MiB) copied, 1.01014 s, 298 MB/s

   The copy keeps running after the signal. Send USR1 again later for another point-in-time report, or use status=progress next time when the command can be restarted safely.

7. Wait for the temporary background copy in the terminal where it was started.

   $ wait %1

8. Remove the temporary sample files.

   $ rm -f /tmp/dd-demo.img /tmp/dd-fallback.img

   Do not remove a real destination file or device after a production dd run. This cleanup applies only to the temporary files created by the sample commands.