Ubunutu 4
Ubunutu 4
Ubunutu 4
To make matters worse, the number and severity of security threats at the edge
are growing rapidly.
In the past, malicious actors have focused more on nding and exploiting
vulnerabilities on servers, laptops, and workstations than on IoT devices. But with
embedded devices becoming more important, proli c, and mass produced, security
researchers and hackers are increasingly turning their attention to the edge.
Hardware integrity
While the security of IoT software is essential, it is also important to consider
the integrity of the edge hardware itself. As devices trend towards increased
computing power and reduced energy consumption, hardware security does not
always keep up, adding a further layer of risk.
Unlike data centre hardware, IoT devices in the eld can be physically accessed by
malicious actors. As such, everything from low-level mainboard components to
debugging interfaces and buses are viable targets, creating an assortment of new
security challenges.
What is MicroK8s?
MicroK8s is a lightweight, CNCF-certi ed, pure upstream Kubernetes distribution
designed not only for clouds and workstations, but also for IoT devices. Created
by Canonical, the company behind Ubuntu, MicroK8s is delivered as a snap – a
containerised software package that bundles Kubernetes together with all of its
dependencies.
4 https://www.iottechnews.com/news/2021/sep/07/kaspersky-attacks-on-iot-devices-double-in-a-year/
4
Most importantly, being a snap, MicroK8s is isolated from underlying systems,
limiting its access to other system services and resources on an IoT device. To truly
address edge security concerns, MicroK8s version 1.25 takes this concept a step
further with strict con nement.
Con nement and permissions are enforced by the Linux kernel security module
AppArmor, alongside other Linux security features. When strictly con ned
MicroK8s is installed, its metadata is examined and used to derive AppArmor
pro les, seccomp lters, device cgroup rules, and traditional permissions.
Together, these ensure total isolation for the Kubernetes runtime.
Perhaps the most important consequence of strict con nement being added to
MicroK8s is that it can now be used with Ubuntu Core.
5
What is Ubuntu Core?
Ubuntu Core is a lean, embedded version of Ubuntu created for the edge.
The main goal of Ubuntu Core is to secure the next generation of IoT devices,
and it achieves this through containerisation. Ubuntu Core itself and all
applications deployed on it are packaged as strictly con ned snaps.
This snap-based paradigm takes the bene ts of strict con nement detailed above
and proliferates them throughout the entire device. All applications are fully
isolated from each other and can only interact with the system through interfaces.
This approach is inherently secure and perfect for IoT devices.
Because all applications running on Ubuntu Core must be strictly con ned, it was
not previously possible to pair it with MicroK8s. Now, Ubuntu Core and MicroK8s
can be combined for a seamless path to secure Kubernetes at scale, optimised for
size, performance, and usability at the edge.
Smart home
Let’s examine a hypothetical application of MicroK8s in a smart home. In this smart
home, the fridge acts as a hub controlling multiple smart devices in the house.
That hub is running Ubuntu Core with MicroK8s.
The workloads on the hub communicate with sensors and other devices around
the house. Critically, these devices must not be in uenced by outside updates that
could cause a malign threat later down the line. All updates must come from a
secure source.
6
One day, an update provides a new type of workload with a new MicroK8s pod
that connects to the TV. It locally stores the latest data coming from the TV on the
fridge database, on the Ubuntu Core le system.
With strict con nement, if the Microk8s pod ever encounters unvalidated or
unsafe data, it can still store it without risk. If it is malware, it cannot reach outside
of its MicroK8s sandbox or compromise any other part of the system.
In this way, strictly con ned MicroK8s lets device manufacturers and smart
home platform providers expand their o�erings and deliver the latest feature
updates with con dence that they will not be putting their customers’ homes
and data at risk.
Smart factory
Edge computing is central to the Industry 4.0 manufacturing revolution.
Interconnected IoT devices capture, share, and process data directly on the factory
oor, enabling smart factories to achieve a signi cant degree of autonomy.
Typically, industrial IoT devices have been limited to a single function, such as
a programmable logic controller (PLC) or an industrial gateway. Snaps enable
manufacturers to take the next step and split IoT devices into di�erent containers,
with multiple functions each sitting in their own snap.
Strict con nement keeps this snap-enabled, software-de ned strategy secure
by ensuring that the various functions on an industrial IoT device will only access
the appropriate, permitted resources and services. For example, take a user who
wants to run AI/ML workloads at the edge to predict machine failure based on
vibrations and enable proactive maintenance before it happens. By using Ubuntu
Core, the user can run these workloads in a snap on the same software-de ned
device that is already in use for monitoring the environment. For instance, the
device may further measure ambient temperature, luminosity and humidity levels
on the factory oor and share those with the IT department.
Without strict con nement, the sensors’ measurements may be exposed to the
AI/ML algorithm. This would result in unbalanced datasets leading to biased
predictions that could ultimately lead to machine downtime. With strictly con ned
MicroK8s, the AI/ML app would only “see” what is in its prede ned, sandboxed
environment, eliminating this risk.
7