To help you get started, we've collected the most common ways that scp is being used within popular public projects.

ManageIQ / integration_tests / utils / composite.py View on Github

allow_agent=False,
            look_for_keys=False,
            gss_auth=False)
    scp = None

    while True:
        source, destination, item, items_done = rc._queue.get()
        destination = local(destination)
        destination_dir = local(destination.dirname)
        destination_dir.ensure(dir=True)
        if not destination.check():
            if scp is None:
                scp = SCPClient(client.get_transport())
            try:
                scp.get(source, destination.strpath)
            except SCPException:
                # remote destination didn't exist
                pass
            except (SSHException, socket.timeout):
                # SSH blew up :(
                rc._queue.put((source, destination, item, items_done))
                rc._queue.task_done()
                continue
        rc._progress_update(item, items_done)
        rc._queue.task_done()

View more ways to use scp

Pure python scp module

The scp.py module uses a paramiko transport to send and receive files via the scp1 protocol. This is the protocol as referenced from the openssh scp program, and has only been tested with this implementation.

Example

from paramiko import SSHClient
from scp import SCPClient

ssh = SSHClient()
ssh.load_system_host_keys()
ssh.connect('example.com')

# SCPCLient takes a paramiko transport as an argument
scp = SCPClient(ssh.get_transport())

scp.put('test.txt', 'test2.txt')
scp.get('test2.txt')

# Uploading the 'test' directory with its content in the
# '/home/user/dump' remote directory
scp.put('test', recursive=True, remote_path='/home/user/dump')

scp.close()

$ md5sum test.txt test2.txt
fc264c65fb17b7db5237cf7ce1780769 test.txt
fc264c65fb17b7db5237cf7ce1780769 test2.txt

Using 'with' keyword

from paramiko import SSHClient
from scp import SCPClient

with SSHClient() as ssh:
    ssh.load_system_host_keys()
    ssh.connect('example.com')

    with SCPClient(ssh.get_transport()) as scp:
        scp.put('test.txt', 'test2.txt')
        scp.get('test2.txt')

$ md5sum test.txt test2.txt
fc264c65fb17b7db5237cf7ce1780769 test.txt
fc264c65fb17b7db5237cf7ce1780769 test2.txt

Uploading file-like objects

The putfo method can be used to upload file-like objects:

import io
from paramiko import SSHClient
from scp import SCPClient

ssh = SSHClient()
ssh.load_system_host_keys()
ssh.connect('example.com')

# SCPCLient takes a paramiko transport as an argument
scp = SCPClient(ssh.get_transport())

# generate in-memory file-like object
fl = io.BytesIO()
fl.write(b'test')
fl.seek(0)
# upload it directly from memory
scp.putfo(fl, '/tmp/test.txt')
# close connection
scp.close()
# close file handler
fl.close()

Tracking progress of your file uploads/downloads

A progress function can be given as a callback to the SCPClient to handle how the current SCP operation handles the progress of the transfers. In the example below we print the percentage complete of the file transfer.

from paramiko import SSHClient
from scp import SCPClient
import sys

ssh = SSHClient()
ssh.load_system_host_keys()
ssh.connect('example.com')

# Define progress callback that prints the current percentage completed for the file
def progress(filename, size, sent):
    sys.stdout.write("%s's progress: %.2f%%   \r" % (filename, float(sent)/float(size)*100) )

# SCPCLient takes a paramiko transport and progress callback as its arguments.
scp = SCPClient(ssh.get_transport(), progress=progress)

# you can also use progress4, which adds a 4th parameter to track IP and port
# useful with multiple threads to track source
def progress4(filename, size, sent, peername):
    sys.stdout.write("(%s:%s) %s's progress: %.2f%%   \r" % (peername[0], peername[1], filename, float(sent)/float(size)*100) )
scp = SCPClient(ssh.get_transport(), progress4=progress4)

scp.put('test.txt', '~/test.txt')
# Should now be printing the current progress of your put function.

scp.close()