Opt-out a bunch of slides in the deep dive section

slides/intro/Namespaces_Cgroups.md

@@ -76,6 +76,8 @@ The last item should be done for educational purposes only!
 
 ---
 
+class: extra-details, deep-dive
+
 ## Manipulating namespaces
 
 - Namespaces are created with two methods:
@@ -94,6 +96,8 @@ The last item should be done for educational purposes only!
 
 ---
 
+class: extra-details, deep-dive
+
 ## Namespaces lifecycle
 
 - When the last process of a namespace exits, the namespace is destroyed.
@@ -114,6 +118,8 @@ The last item should be done for educational purposes only!
 
 ---
 
+class: extra-details, deep-dive
+
 ## Namespaces can be used independently
 
 - As mentioned in the previous slides:
@@ -150,6 +156,8 @@ The last item should be done for educational purposes only!
 
 ---
 
+class: extra-details, deep-dive
+
 ## Creating our first namespace
 
 Let's use `unshare` to create a new process that will have its own UTS namespace:
@@ -166,6 +174,8 @@ $ sudo unshare --uts
 
 ---
 
+class: extra-details, deep-dive
+
 ## Demonstrating our uts namespace
 
 In our new "container", check the hostname, change it, and check it:
@@ -398,6 +408,8 @@ class: extra-details
 
 ---
 
+class: extra-details, deep-dive
+
 ## Setting up a private `/tmp`
 
 Create a new mount namespace:
@@ -435,6 +447,8 @@ The mount is automatically cleaned up when you exit the process.
 
 ---
 
+class: extra-details, deep-dive
+
 ## PID namespace in action
 
 Create a new PID namespace:
@@ -453,10 +467,14 @@ Check the process tree in the new namespace:
 
 --
 
+class: extra-details, deep-dive
+
 🤔 Why do we see all the processes?!?
 
 ---
 
+class: extra-details, deep-dive
+
 ## PID namespaces and `/proc`
 
 - Tools like `ps` rely on the `/proc` pseudo-filesystem.
@@ -471,6 +489,8 @@ Check the process tree in the new namespace:
 
 ---
 
+class: extra-details, deep-dive
+
 ## PID namespaces, take 2
 
 - This can be solved by mounting `/proc` in the namespace.
@@ -570,6 +590,8 @@ Check `man 2 unshare` and `man pid_namespaces` if you want more details.
 
 ---
 
+class: extra-details, deep-dive
+
 ## User namespace challenges
 
 - UID needs to be mapped when passed between processes or kernel subsystems.
@@ -686,6 +708,8 @@ cpu                      memory
 
 ---
 
+class: extra-details, deep-dive
+
 ## Cgroups v1 vs v2
 
 - Cgroups v1 are available on all systems (and widely used).
@@ -759,6 +783,8 @@ cpu                      memory
 
 ---
 
+class: extra-details, deep-dive
+
 ## Avoiding the OOM killer
 
 - For some workloads (databases and stateful systems), killing
@@ -778,6 +804,8 @@ cpu                      memory
 
 ---
 
+class: extra-details, deep-dive
+
 ## Overhead of the memory cgroup
 
 - Each time a process grabs or releases a page, the kernel update counters.
@@ -796,6 +824,8 @@ cpu                      memory
 
 ---
 
+class: extra-details, deep-dive
+
 ## Setting up a limit with the memory cgroup
 
 Create a new memory cgroup:
@@ -823,6 +853,8 @@ The current process *and all its future children* are now limited.
 
 ---
 
+class: extra-details, deep-dive
+
 ## What's `<<<`?
 
 - This is a "here string". (It is a non-POSIX shell extension.)
@@ -847,6 +879,8 @@ The current process *and all its future children* are now limited.
 
 ---
 
+class: extra-details, deep-dive
+
 ## Writing to cgroups pseudo-files requires root
 
 Instead of:
@@ -874,6 +908,8 @@ echo $$ > $CG/tasks
 
 ---
 
+class: extra-details, deep-dive
+
 ## Testing the memory limit
 
 Start the Python interpreter: