Added a paragraph about Baleine for basic explanation and link to the Baleine wiki.

.gitignore

@@ -3,3 +3,6 @@ venv/
 assets/unwrap/
 *.dot
 manifold/
+RANDOM
+db.sqlite3
+django.log

code/A5-ping.py

@@ -47,30 +47,15 @@
     scheduler = scheduler,
 )
 
-##########
-# setting up the data interface on both fit01 and fit02
-init_node_01 = SshJob(
-    node = node1,
-    command = Run("turn-on-data"),
-    required = check_lease,
-    scheduler = scheduler,
-)
-init_node_02 = SshJob(
-    node = node2,
-    command = Run("turn-on-data"),
-    required = check_lease,
-    scheduler = scheduler,
-)
-
 # the command we want to run in node1 is as simple as it gets
 ping = SshJob(
     node = node1,
     # wait for the 2 init jobs instead
     # check_release is guaranteed to have completed anyway
-    required = (init_node_01, init_node_02),
+    required = (check_lease),
     # let's be more specific about what to run
     # we will soon see other things we can do on an ssh connection
-    command = Run('ping', '-c1', '-I', 'data', 'data02'),
+    command = Run('ping', '-c1', 'fit02'),
     scheduler = scheduler,
 )
 

code/C1-files.py

@@ -37,7 +37,7 @@
                   verbose = verbose_ssh)
 
 # saying gateway = faraday means to tunnel ssh through the gateway
-node1 = SshNode(gateway = faraday, hostname = "fit01", username = "root",
+node1 = SshNode(gateway = faraday, hostname = "fit20", username = "root", port = 2222,
                 verbose = verbose_ssh)
 
 ########## create the scheduler instance upfront

markdown/tuto-020-shell-tools.md

@@ -7,7 +7,7 @@ skip_header: True
 <script src="/assets/r2lab/r2lab-diff.js"></script>
 <style>@import url("/assets/r2lab/r2lab-diff.css")</style>
 
-<< tuto_tabs "LOG IN":LOGIN  "SELECT NODES":NODES "IMAGES": "PHONES": >>
+<< tuto_tabs "LOG IN":LOGIN  "SELECT NODES":NODES "IMAGES": "PHONES": "BALEINE": >>
 
 <div id="contents" class="tab-content" markdown="1">
 
@@ -228,6 +228,13 @@ This command, like all the `rhubarbe`-related commands, has a default timeout, t
 
     rwait --timeout 30
 
+### Using Docker containers with baleine
+
+Baleine is a CLI tool allowing you to deploy Docker images to nodes running on the R2Lab testbed.
+The default disk image to use with all the Docker configuration already done is the `baleine` image.
+
+If you want exhaustive information about the different options available in the `baleine` CLI, please check out the [official docs](https://github.com/haysberg/baleine/wiki) on GitHub.
+
 
 ### `ssh`-ing into nodes
 
@@ -249,6 +256,10 @@ You can run a command on all selected nodes with
 
 this time of course, you cannot specify another set of nodes than the selection.
 
+If you want to directly access a Bash prompt **inside a Docker container** please connect through port `2222`, like this :
+
+    ssh -p 2222 root@fit25
+
 ### Saving images
 
 You have the ablility to save the image - this now of course applies only to **one node** at a time. To save node 25
@@ -287,4 +298,34 @@ to get a reminder.
 
 </div>
 
+<!-- ------- BALEINE ------------>
+<div id="BALEINE" class="tab-pane fade" markdown="1">
+
+## Baleine
+
+Baleine is a gateway tool allowing you to deploy Docker containers inside the R2Lab testbed.
+
+We will now demonstrate the deploy subcommand that will be very useful to you :
+
+```
+baleine deploy --image faraday.repo/tutorial --nodes 1 2 --options -t -d
+```
+
+The `deploy` subcommand allows you to pull and deploy the docker image selected with `--image` on the nodes selected with `--nodes`.
+The `--options` option allows you to give an array of argument to pass directly to the Docker Runtime.
+
+The `faraday.repo/tutorial` docker image is based on Ubuntu, so the `-t` and `-d` options are necessary for the Docker image to continue running after initial launch.
+
+To use the Docker container as your server to run the tutorials, please use port `2222` for SSH as `22` is reserved for the host OS.
+
+### Saving Docker images
+
+If you want to save a Docker container that you may have modified from its base image, you can use the [baleine save](https://github.com/haysberg/baleine/wiki/Save-a-custom-container) command.
+
+    baleine save --node 1 --name mycustomimage:1.0 
+
+[Check out the wiki](https://github.com/haysberg/baleine/wiki) for exhaustive information on the different commands and options available.
+
+</div>
+
 </div> <!-- end div contents -->

markdown/tuto-040-A-ping.md

@@ -36,9 +36,10 @@ in each tutorial.
 Before we can run these experiments however, we need to have
 
 * a valid lease set in the booking system
-* the 2 nodes `fit01` and `fit02` up and running
+* the 2 nodes `fit01` and `fit02` up and running with the `baleine` disk image
+* the `faraday.repo/tutorial` Docker image running on the fit nodes
 
-For this first tutorial we will assume that these 2 steps have been
+For this first tutorial we will assume that these 3 steps have been
 performed manually, and here is how to proceed with that.
 
 ### Getting a reservation
@@ -60,7 +61,7 @@ The code in this tutorial assumes you have a slice named
 replace with your actual slice name when trying to run the code
 yourself.
 
-### Loading images
+### Loading disk images
 
 For loading the images manually on the 2 nodes needed here, please do
 this (as usual, make sure to use **your slice name** instead of
@@ -86,7 +87,7 @@ explained [in the previous tutorial](tuto-020-shell-tools.md):
     # select nodes 1 and 2
     n 1 2
     # load the default image (on the selected nodes)
-    rload
+    rload -i baleine
     # wait for ssh to be up (still on the selected nodes)
     rwait
 
@@ -119,27 +120,49 @@ should look like
     export NBNODES=2
 
     your_slicename@faraday:~$ rload
-    16:12:42 - +000s: Selection: fit01 fit02
-    16:12:42 - +000s: Loading image /var/lib/rhubarbe-images/default.ndz
-    16:12:42 - +000s: AUTH: checking for a valid lease
-    16:12:42 - +000s: AUTH: access granted
-    16:12:42 - +000s: fit02 reboot = Sending message 'on' to CMC reboot02
-    16:12:42 - +000s: fit01 reboot = Sending message 'on' to CMC reboot01
-    16:12:43 - +001s: fit02 reboot = idling for 15s
-    16:12:43 - +001s: fit01 reboot = idling for 15s
-    16:12:59 - +017s: started <[email protected]:10001 on default.ndz at 500 Mibps>
-    16:12:59 - +017s: fit01 frisbee_status = trying to telnet..
-    16:12:59 - +017s: fit02 frisbee_status = trying to telnet..
-    ... <snip>
-    |############################################################################################|100% |29.56s|Time: 0:00:29
-    16:13:44 - +062s: fit02 Uploading successful
-    16:13:44 - +062s: fit02 reboot = Sending message 'reset' to CMC reboot02
-    16:13:46 - +064s: stopped <[email protected]:10001 on default.ndz at 500 Mibps>
+    Found binary frisbeed as /usr/sbin/frisbeed
+    Found binary nc as /usr/bin/nc
+    09:38:23 - +000s: Selection: fit01 fit02
+    09:38:23 - +000s: Loading image /var/lib/rhubarbe-images/baleine.ndz
+    09:38:23 - +000s: AUTH: checking for a valid lease
+    09:38:23 - +000s: AUTH: access granted
+    09:38:23 - +000s: fit01 reboot = Sending message 'on' to CMC reboot01
+    09:38:23 - +000s: fit02 reboot = Sending message 'on' to CMC reboot02
+    09:38:24 - +001s: fit02 reboot = idling for 30s
+    09:38:24 - +001s: fit01 reboot = idling for 30s
+    09:38:55 - +032s: started <[email protected]:10001 on baleine.ndz at 500 Mibps>
+    09:38:55 - +032s: fit01 frisbee_status = trying to telnet..
+    09:38:55 - +032s: fit02 frisbee_status = trying to telnet..
+    09:38:56 - +032s: fit01 frisbee_status = backing off for 4.22s
+    09:38:56 - +032s: fit02 frisbee_status = backing off for 3.07s
+    09:38:59 - +035s: fit02 frisbee_status = trying to telnet..
+    09:38:59 - +036s: fit02 frisbee_status = backing off for 1.61s
+    09:39:00 - +036s: fit01 frisbee_status = trying to telnet..
+    09:39:00 - +037s: fit01 frisbee_status = starting frisbee client
+    09:39:01 - +037s: fit02 frisbee_status = trying to telnet..                  |  0% |0.00s|ETA:  --:--:--
+    09:39:01 - +038s: fit02 frisbee_status = starting frisbee client
+    09:40:00 - +096s: fit02 reboot = Sending message 'reset' to CMC reboot02#### | 99% |59.46s|ETA:  0:00:00
+    |############################################################################|100% |59.81s|Time: 0:00:59
+    09:40:00 - +097s: fit01 reboot = Sending message 'reset' to CMC reboot01
+    09:40:02 - +099s: stopped <[email protected]:10001 on baleine.ndz at 500 Mibps>
 
     your_slicename@faraday:~$ rwait
     <Node fit01>:ssh OK
     <Node fit02>:ssh OK
 
+### Loading the Tutorial Docker image
+
+To load Docker image on the fit nodes, we will now use the `baleine` command :
+
+    baleine deploy --nodes 1 2 --image faraday.repo/tutorial --options -t -d
+
+Let's give a brief explanation of what we just wrote :
+* The `deploy` subcommand allows us to pull an image from a Docker repository and create/launch a container on a fit node.
+* The `--nodes` option allows us to mention what nodes we want to target. You can format the nodes just like with the `nodes` command you used before.
+* The `--image` argument is the url to pull the image from. This can be from the Faraday repository, GitHub, DockerHub or any public repository. Please not that `faraday.repo` is reachable **ONLY** from the fit nodes.
+* The `--options` argument allows to pass specific arguments directly to the Docker daemon running on the fit nodes. By default, if you give no command the `ubuntu` Docker image will stop instantly. To prevent that, we use the `-t -d` options. You can get more details about them [here](https://docs.docker.com/engine/reference/run/). 
+
+
 At this point, both nodes have been loaded with the default image. So
 you can log out of `faraday.inria.fr` and go back to your laptop to
 run [the tutorial in tab A1](javascript:open_tab('A1')).
@@ -257,11 +280,13 @@ This materializes the fact that we reach node `fit01` through the
 gateway. It also ensures that only one ssh connection gets established
 to the gateway, regardless of the number of nodes actually controlled.
 
+Please note that we are using the `container` user in the script to directly connect to the Docker container shell.
+
 **Double check:** Remember that for this to work, you need to have a currently valid slice (and to use it with the `--slice` option if needed), and you need `fit01` to be up and running.
 
 ### The code
 
-<< codeview A3 A3-ping.py previous=A2-ping.py >>
+<< codeview A3 A3-ping-docker.py previous=A2-ping.py >>
 
 ### Sample output
 
@@ -349,7 +374,7 @@ for an example of that feature.
 
 ### The code
 
-<< codeview A4 A4-ping.py previous=A3-ping.py graph=A4.png >>
+<< codeview A4 A4-ping-docker.py previous=A3-ping-docker.py graph=A4.png >>
 
 ### Sample output
 

r2lab/settings.py

@@ -68,6 +68,7 @@
 ALLOWED_HOSTS = [
     'r2lab.inria.fr',
     'localhost',
+    '127.0.0.1'
 ]
 
 # Application definition