The ssh Package: Secure Shell (SSH) Client for R

[This article was first published on rOpenSci - open tools for open science, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

cover image

Have you ever needed to connect to a remote server over SSH to transfer files via SCP or to setup a secure tunnel, and wished you could do so from R itself? The new rOpenSci ssh package provides a native ssh client in R allows you to do that and even more, like running a command or script on the host while streaming stdout and stderr directly to the client. The package is based on libssh, a powerful C library implementing the SSH protocol.

install.packages("ssh")

Because the ssh package is based on libssh it does not need to shell out. Therefore it works natively on all platforms without any runtime dependencies. Even on Windows.

The package is still work in progress, but the core functionality should work. Below some examples to get you started from the intro vignette.

Connecting to an SSH server

First create an ssh session by connecting to an SSH server.

session <- ssh_connect("[email protected]")
print(session)

## <ssh session>
## connected: [email protected]:22
## server: 1e:28:44:af:84:91:e5:88:fe:82:ca:34:d7:c8:cf:a8:0d:2f:ec:af

Once established, a session is closed automatically by the garbage collector when the object goes out of scope or when R quits. You can also manually close it using ssh_disconnect() but this is not strictly needed.

Authentication

The client attempts to use the following authentication methods (in this order) until one succeeds:

  1. try key from privkey argument in ssh_connect() if specified
  2. if ssh-agent is available, try private key from ssh-agent
  3. try user key specified in ~/.ssh/config or any of the default locations: ~/.ssh/id_ed25519, ~/.ssh/id_ecdsa, ~/.ssh/id_rsa, or .ssh/id_dsa.
  4. Try challenge-response password authentication (if permitted by the server)
  5. Try plain password authentication (if permitted by the server)

To debug authentication set verbosity to at least level 2 or 3:

session <- ssh_connect("[email protected]", verbose = 2)
## ssh_socket_connect: Nonblocking connection socket: 7
## ssh_connect: Socket connecting, now waiting for the callbacks to work
## socket_callback_connected: Socket connection callback: 1 (0)
## ssh_client_connection_callback: SSH server banner: SSH-2.0-OpenSSH_7.2p2 Ubuntu-4ubuntu2.4
## ssh_analyze_banner: Analyzing banner: SSH-2.0-OpenSSH_7.2p2 Ubuntu-4ubuntu2.4
## ssh_analyze_banner: We are talking to an OpenSSH client version: 7.2 (70200)
## ssh_packet_dh_reply: Received SSH_KEXDH_REPLY
## ssh_client_curve25519_reply: SSH_MSG_NEWKEYS sent
## ssh_packet_newkeys: Received SSH_MSG_NEWKEYS
## ssh_packet_newkeys: Signature verified and valid
## ssh_packet_userauth_failure: Access denied. Authentication that can continue: publickey
## ssh_packet_userauth_failure: Access denied. Authentication that can continue: publickey
## ssh_agent_get_ident_count: Answer type: 12, expected answer: 12
## ssh_userauth_publickey_auto: Successfully authenticated using /Users/jeroen/.ssh/id_rsa

Execute Script or Command

Run a command or script on the host and block while it runs. By default stdout and stderr are steamed directly back to the client. This function returns the exit status of the remote command (hence it does not automatically error for an unsuccessful exit status).

out <- ssh_exec_wait(session, command = 'whoami')
##  jeroen

print(out)
##  [1] 0

You can also run a script that consists of multiple commands.

ssh_exec_wait(session, command = c(
  'curl -O https://cran.r-project.org/src/contrib/Archive/jsonlite/jsonlite_1.4.tar.gz',
  'R CMD check jsonlite_1.4.tar.gz',
  'rm -f jsonlite_1.4.tar.gz'
))
##   % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
##                                  Dload  Upload   Total   Spent    Left  Speed
## 100 1071k  100 1071k    0     0   654k      0  0:00:01  0:00:01 --:--:--  654k
## * using log directory '/home/jeroen/jsonlite.Rcheck'
## * using R version 3.4.3 (2017-11-30)
## * using platform: x86_64-pc-linux-gnu (64-bit)
## * using session charset: ASCII
## * checking for file 'jsonlite/DESCRIPTION' ... OK
## * this is package 'jsonlite' version '1.4'
## * checking package namespace information ... OK
## * checking package dependencies ...
...

Capturing output

The ssh_exec_internal() is a convenient wrapper for ssh_exec_wait() which buffers the output steams and returns them as a raw vector. Also it raises an error by default when the remote command was not successful.

out <- ssh_exec_internal(session, "R -e 'rnorm(10)'")
print(out$status)
##  [1] 0

cat(rawToChar(out$stdout))

## R version 3.4.4 (2018-03-15) -- "Someone to Lean On"
## Copyright (C) 2018 The R Foundation for Statistical Computing
## Platform: x86_64-pc-linux-gnu (64-bit)
## 
## R is free software and comes with ABSOLUTELY NO WARRANTY.
## You are welcome to redistribute it under certain conditions.
## Type 'license()' or 'licence()' for distribution details.
## 
## R is a collaborative project with many contributors.
## Type 'contributors()' for more information and
## 'citation()' on how to cite R or R packages in publications.
## 
## Type 'demo()' for some demos, 'help()' for on-line help, or
## 'help.start()' for an HTML browser interface to help.
## Type 'q()' to quit R.
## 
## > rnorm(10)
##  [1]  0.14301778 -0.26873489  0.83931307  0.22034917  0.87214122 -0.13655736
##  [7] -0.08793867 -0.68616146  0.23469591  0.93871035

This function is very useful if you are running a remote command and want to use it’s output as if you had executed it locally.

Using sudo

Note that the exec functions are non interactive so they cannot prompt for a sudo password. A trick is to use -S which reads the password from stdin:

command <- 'echo "mypassword!" | sudo -s -S apt-get update -y'
out <- ssh_exec_wait(session, command)

Be very careful with hardcoding passwords!

Transfer Files via SCP

Upload and download files via SCP. Directories are automatically traversed as in scp -r.

# Upload a file to the server
file_path <- R.home("COPYING")
scp_upload(session, file_path)
## [100%] /Library/Frameworks/R.framework/Versions/3.5/Resources/COPYING

This will upload the file to the home directory on your server. Let’s download it back:

# Download the file back and verify it is the same
scp_download(session, "COPYING", to = tempdir())
## 18011 /var/folders/l8/bhmtp25n2lx0q0dgv1x4gf1w0000gn/T//Rtmpldz4eO/COPYING

We can compare the checksums to verify that the files are identical:

tools::md5sum(file_path)
##  "eb723b61539feef013de476e68b5c50a" 
tools::md5sum(file.path(tempdir(), "COPYING"))
##  "eb723b61539feef013de476e68b5c50a" 

Hosting a Tunnel

Opens a port on your machine and tunnel all traffic to a custom target host via the SSH server.

ssh_tunnel(session, port = 5555, target = "ds043942.mongolab.com:43942")

This function blocks while the tunnel is active. Use the tunnel by connecting to localhost:5555 from a separate process. The tunnel can only be used once and will automatically be closed when the client disconnects.

To leave a comment for the author, please follow the link and comment on their blog: rOpenSci - open tools for open science.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)