diff --git a/public/tuto2/index.html b/public/tuto2/index.html
index daed5641157b88a5364aa4f48bca9bc3ef1a9364..de41c546b93c47ae591499618d51680755067fe8 100644
--- a/public/tuto2/index.html
+++ b/public/tuto2/index.html
@@ -3,7 +3,7 @@
 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
 <html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">
 <head>
-<!-- 2019-11-29 ven. 00:21 -->
+<!-- 2019-11-29 ven. 01:19 -->
 <meta http-equiv="Content-Type" content="text/html;charset=utf-8" />
 <meta name="viewport" content="width=device-width, initial-scale=1" />
 <title>Distributed experiments on Grid'5000 &#x2026; and beyond !</title>
@@ -235,17 +235,24 @@ for the JavaScript code in this tag.
 <h2>Table of Contents</h2>
 <div id="text-table-of-contents">
 <ul>
-<li><a href="#orgc38ce74">1. Benchmarking a real application</a></li>
-<li><a href="#org3f66fb1">2. Before you start</a>
+<li><a href="#org9bd26d8">1. Benchmarking a real application</a></li>
+<li><a href="#org8d16b67">2. Before you start</a>
 <ul>
-<li><a href="#orga16f997">2.1. Grid'5000 stuffs</a></li>
-<li><a href="#orgfd9377b">2.2. Setup on Grid'5000</a></li>
+<li><a href="#org3e3133c">2.1. Grid'5000 stuffs</a></li>
+<li><a href="#org0f73ed8">2.2. Setup on Grid'5000</a></li>
 </ul>
 </li>
-<li><a href="#orgc0bc76f">3. Deployment time !</a>
+<li><a href="#orgb804170">3. Deployment time !</a>
 <ul>
-<li><a href="#org1df303b">3.1. Deploy it</a></li>
-<li><a href="#org8a7d969">3.2. Access it</a></li>
+<li><a href="#org53db06f">3.1. Deploy it</a></li>
+<li><a href="#org4a87400">3.2. Access it</a></li>
+</ul>
+</li>
+<li><a href="#org51cf531">4. Deploy the monitoring stack</a></li>
+<li><a href="#orga6debd5">5. Benchmark the system</a>
+<ul>
+<li><a href="#org04ec531">5.1. Deploy the benchmarking nodes</a></li>
+<li><a href="#orgc0a0755">5.2. Observations</a></li>
 </ul>
 </li>
 </ul>
@@ -253,18 +260,18 @@ for the JavaScript code in this tag.
 </div>
 
 
-<div id="outline-container-orgc38ce74" class="outline-2">
-<h2 id="orgc38ce74"><span class="section-number-2">1</span> Benchmarking a real application</h2>
+<div id="outline-container-org9bd26d8" class="outline-2">
+<h2 id="org9bd26d8"><span class="section-number-2">1</span> Benchmarking a real application</h2>
 <div class="outline-text-2" id="text-1">
 <p>
 In this tutorial we'll cover some aspects of evaluating the performance of a
 real application. We'll work with <code>overleaf</code>. <code>overleaf</code> is a collaborative
-text editor that uses Latex to produce pdf files. Figure <a href="#org3218667">1</a>
+text editor that uses Latex to produce pdf files. Figure <a href="#org822339d">1</a>
 is an overview of the editing part of the software.
 </p>
 
 
-<div id="org3218667" class="figure">
+<div id="org822339d" class="figure">
 <p><a href="./figs/overleaf-v2-editor.png"><img src="./figs/overleaf-v2-editor.png" alt="overleaf-v2-editor.png" /></a>
 </p>
 <p><span class="figure-number">Figure 1: </span>Overview of <code>overleaf</code> editor: on the left users can collaboratively edit the document. On the right the document is rendered.</p>
@@ -285,16 +292,16 @@ the official/commercial instance).</li>
 </div>
 </div>
 
-<div id="outline-container-org3f66fb1" class="outline-2">
-<h2 id="org3f66fb1"><span class="section-number-2">2</span> Before you start</h2>
+<div id="outline-container-org8d16b67" class="outline-2">
+<h2 id="org8d16b67"><span class="section-number-2">2</span> Before you start</h2>
 <div class="outline-text-2" id="text-2">
 <p>
 Make sure you are ok with the following.
 </p>
 </div>
 
-<div id="outline-container-orga16f997" class="outline-3">
-<h3 id="orga16f997"><span class="section-number-3">2.1</span> Grid'5000 stuffs</h3>
+<div id="outline-container-org3e3133c" class="outline-3">
+<h3 id="org3e3133c"><span class="section-number-3">2.1</span> Grid'5000 stuffs</h3>
 <div class="outline-text-3" id="text-2-1">
 <div class="note">
 <p>
@@ -308,8 +315,8 @@ we'll prefer to use higher level tools for now.
 </div>
 
 
-<div id="outline-container-orgfd9377b" class="outline-3">
-<h3 id="orgfd9377b"><span class="section-number-3">2.2</span> Setup on Grid'5000</h3>
+<div id="outline-container-org0f73ed8" class="outline-3">
+<h3 id="org0f73ed8"><span class="section-number-3">2.2</span> Setup on Grid'5000</h3>
 <div class="outline-text-3" id="text-2-2">
 <p>
 Connect to a Grid'5000 frontend of your choice (e.g rennes, nancy &#x2026;)
@@ -336,11 +343,11 @@ $<span style="color: #7590db;">frontend</span><span style="color: #4f97d7;">(</s
 </div>
 </div>
 
-<div id="outline-container-orgc0bc76f" class="outline-2">
-<h2 id="orgc0bc76f"><span class="section-number-2">3</span> Deployment time !</h2>
+<div id="outline-container-orgb804170" class="outline-2">
+<h2 id="orgb804170"><span class="section-number-2">3</span> Deployment time !</h2>
 <div class="outline-text-2" id="text-3">
 <p>
-Figure <a href="#org4b6858a">2</a> represents a simplified view of what we'll deploy. In blue
+Figure <a href="#orgbf401ab">2</a> represents a simplified view of what we'll deploy. In blue
 some services of <code>overleaf</code> are represented. First the Web portal is the entry
 point to all the user requests. The three other services in the picture are
 involved when compiling a document. The compilation service interacts with the
@@ -351,7 +358,7 @@ involved to provide the chat feature, history feature, real-time interaction
 </p>
 
 
-<div id="org4b6858a" class="figure">
+<div id="orgbf401ab" class="figure">
 <p><a href="./figs/simple_compilation.png"><object type="image/svg+xml" data="./figs/simple_compilation.svg" class="org-svg">
 Sorry, your browser does not support SVG.</object></a>
 </p>
@@ -371,8 +378,8 @@ After the deployment you'll be able to:
 </div>
 </div>
 
-<div id="outline-container-org1df303b" class="outline-3">
-<h3 id="org1df303b"><span class="section-number-3">3.1</span> Deploy it</h3>
+<div id="outline-container-org53db06f" class="outline-3">
+<h3 id="org53db06f"><span class="section-number-3">3.1</span> Deploy it</h3>
 <div class="outline-text-3" id="text-3-1">
 <div class="org-src-container">
 <pre class="src src-bash">$<span style="color: #7590db;">frontend</span><span style="color: #4f97d7;">(</span>venv<span style="color: #4f97d7;">)</span>: python overleaf.py deploy --cluster=paravance
@@ -389,8 +396,8 @@ After the deployment you'll be able to:
 </div>
 </div>
 
-<div id="outline-container-org8a7d969" class="outline-3">
-<h3 id="org8a7d969"><span class="section-number-3">3.2</span> Access it</h3>
+<div id="outline-container-org4a87400" class="outline-3">
+<h3 id="org4a87400"><span class="section-number-3">3.2</span> Access it</h3>
 <div class="outline-text-3" id="text-3-2">
 <p>
 To know where your services is installed you can run: 
@@ -416,7 +423,7 @@ To know where your services is installed you can run:
 
 <div class="note">
 <p>
-To access the web portal you can create a tunnel from your local machine to
+To access the web portal, you can create a tunnel from your local machine to
 the machine running the web portal as follows
 </p>
 
@@ -430,6 +437,97 @@ $<span style="color: #7590db;">yourmachine</span>: ssh -NL <span style="color: #
 </pre>
 </div>
 
+</div>
+
+<div class="note">
+<p>
+You can access the web portal of your friends by replacing the address of the web portal.
+</p>
+
+</div>
+</div>
+</div>
+</div>
+
+<div id="outline-container-org51cf531" class="outline-2">
+<h2 id="org51cf531"><span class="section-number-2">4</span> Deploy the monitoring stack</h2>
+<div class="outline-text-2" id="text-4">
+<div class="org-src-container">
+<pre class="src src-bash">$<span style="color: #7590db;">frontend</span><span style="color: #4f97d7;">(</span>venv<span style="color: #4f97d7;">)</span>: python overleaf.py monitoring
+</pre>
+</div>
+
+<div class="question">
+<ul class="org-ul">
+<li>Create another tunnel and access the monitoring portal</li>
+<li>Import the dashboard [TODO]</li>
+</ul>
+
+</div>
+</div>
+</div>
+
+<div id="outline-container-orga6debd5" class="outline-2">
+<h2 id="orga6debd5"><span class="section-number-2">5</span> Benchmark the system</h2>
+<div class="outline-text-2" id="text-5">
+<p>
+Benchmarking will consist in generating some load on the system. 
+This will be made in two steps:
+</p>
+<ul class="org-ul">
+<li>First, we'll install a benchmarking portal and some agent responsible
+for generating the load</li>
+<li>Second, from the portal we'll generate some load</li>
+</ul>
+
+<p>
+The load will consist in (many) users creating a project and compiling it once
+before destroying it
+</p>
+</div>
+
+<div id="outline-container-org04ec531" class="outline-3">
+<h3 id="org04ec531"><span class="section-number-3">5.1</span> Deploy the benchmarking nodes</h3>
+<div class="outline-text-3" id="text-5-1">
+<div class="org-src-container">
+<pre class="src src-bash">$<span style="color: #7590db;">frontend</span><span style="color: #4f97d7;">(</span>venv<span style="color: #4f97d7;">)</span>: python overleaf.py bench
+</pre>
+</div>
+
+<div class="question">
+<ul class="org-ul">
+<li>Create another tunnel and access the benchmarking portal</li>
+<li>Generate a small load (1 user)</li>
+</ul>
+
+</div>
+</div>
+</div>
+
+<div id="outline-container-orgc0a0755" class="outline-3">
+<h3 id="orgc0a0755"><span class="section-number-3">5.2</span> Observations</h3>
+<div class="outline-text-3" id="text-5-2">
+<p>
+Observations can be made on two portals:
+</p>
+<ul class="org-ul">
+<li>the benchmarking portal: you'll get some information on the requests sent
+to the server</li>
+<li>the monitorig portal: you'll get some information about the resource
+consumption and some application metrics</li>
+</ul>
+
+<div class="question">
+<ul class="org-ul">
+<li>Can you see the effect of the load ?</li>
+<li>Scale the workload by starting 10 users
+<ul class="org-ul">
+<li>what kind of metrics seem impacted on the monitoring portal ?</li>
+</ul></li>
+<li>the projects that are compiled aren't all identical, can you characterize them ?
+(in term of CPU/IO &#x2026; consumed)</li>
+</ul>
+
 </div>
 </div>
 </div>
@@ -437,7 +535,7 @@ $<span style="color: #7590db;">yourmachine</span>: ssh -NL <span style="color: #
 </div>
 <div id="postamble" class="status">
 <p class="author">Author: Matthieu Simonin</p>
-<p class="date">Created: 2019-11-29 ven. 00:21</p>
+<p class="date">Created: 2019-11-29 ven. 01:19</p>
 <p class="validation"><a href="http://validator.w3.org/check?uri=referer">Validate</a></p>
 </div>
 </body>
diff --git a/public/tuto2/index.org b/public/tuto2/index.org
index f915d1e07a48fc6fd0e05922ea0b8fe845d5a6f1..7f360c29dc257824d078e88b6607ccc46911cf13 100644
--- a/public/tuto2/index.org
+++ b/public/tuto2/index.org
@@ -124,7 +124,7 @@
   #+END_SRC
 
   #+BEGIN_note
-  To access the web portal you can create a tunnel from your local machine to
+  To access the web portal, you can create a tunnel from your local machine to
   the machine running the web portal as follows
 
   #+BEGIN_SRC bash :noeval
@@ -137,11 +137,56 @@
   #+END_SRC
   #+END_note
 
+  #+BEGIN_note
+  You can access the web portal of your friends by replacing the address of the web portal.
+  #+END_note
+
+* Deploy the monitoring stack
+  
+  #+BEGIN_SRC bash :noeval
+  $frontend(venv): python overleaf.py monitoring
+  #+END_SRC
+
+  #+begin_question
+  - Create another tunnel and access the monitoring portal
+  - Import the dashboard [TODO]
+  #+end_question
 
 * Benchmark the system
 
-** Deploy the benchmark nodes
-** Prepare the monitoring stack
+  Benchmarking will consist in generating some load on the system. 
+  This will be made in two steps:
+  - First, we'll install a benchmarking portal and some agent responsible
+    for generating the load
+  - Second, from the portal we'll generate some load
+
+  The load will consist in (many) users creating a project and compiling it once
+  before destroying it
+
+** Deploy the benchmarking nodes
+
+  #+BEGIN_SRC bash :noeval
+  $frontend(venv): python overleaf.py bench
+  #+END_SRC
+
+  #+begin_question
+  - Create another tunnel and access the benchmarking portal
+  - Generate a small load (1 user)
+  #+end_question
+
 ** Observations
+   
+   Observations can be made on two portals:
+   - the benchmarking portal: you'll get some information on the requests sent
+     to the server
+   - the monitorig portal: you'll get some information about the resource
+     consumption and some application metrics
+
+  #+begin_question
+  - Can you see the effect of the load ? 
+  - Scale the workload by starting 10 users
+    + what kind of metrics seem impacted on the monitoring portal ?
+  - the projects that are compiled aren't all identical, can you characterize them ?
+    (in term of CPU/IO ... consumed)
+  #+end_question
 
-* Scale the system