Cloud Containers

Recommended for You
Hyperconverged Infrastructure for Dummies
Hyperconverged Infrastructure for Dummies
Powerful IT doesn’t have to be complicated. Learn how hyperconvergence combines all of your virtualised infrastructure and data services into a single, integrated powerhouse.
Get HCI Dummies Guide

What are cloud containers?

Cloud containers are portable digital compartments holding a bundle of application files in one runtime environment that live in the cloud. They are identical to other technological containers, with the only point of difference being where they are housed.

What do cloud containers do?

Cloud containers are designed to virtualise a single application that is hosted on the cloud. With each container holding just the one application, DevOps can adjust various features as needed without affecting the entire application. This helps accelerate production, allowing for efficient application refinement and rapid deployment at scale.

Related HPE Solutions, Products or Services

How are cloud containers different from virtual machines (VMs)?

The main difference between cloud containers and VMs lies in the level of resources that are dedicated in each. While VMs need a full OS and a virtual copy of the host server’s hardware installed, cloud containers do not. With just a minimum amount of resources, cloud containers can still accomplish what they’re designed for and can be spun up much faster than VMs.

Cloud containers and security

Cloud containers have become a hot topic in the IT industry as cyberattacks persist and major organisations see their platforms fail. Because they offer a measure of protection for IT infrastructure, cloud containers are a popular way for DevOps to continue their production environment on the cloud without exposing their applications to hacking. That’s why top technology companies that have experienced outages, such as Facebook and Instagram, all use them.

How do cloud containers work?

Cloud containers operate in the same way as traditional containers. They virtualise the underlying OS and allow containerised apps to function as if they have a dedicated OS for themselves, including CPU, memory, file storage and network connections. Essentially, the cloud container provides an immutable, lightweight infrastructure for each application that is packaged with its configuration, library requirements and dependencies. These together become a container image hosted on the cloud.

Each one of these container image files are complete and executable, which is accomplished by the container engine. The host OS, however, limits any single container’s access to its physical resources to prevent it from depleting all resources.

IT teams use cloud containers to deploy and run applications in virtual isolation from other applications sharing the same OS kernel. But the containers themselves share the machine OS kernel, which makes their files small and resource-light. It also means that just one singular OS can run many isolated containers. In fact, containers carry all their dependencies with them, enabling deployment without re-configuration for varying environments, such as laptops, cloud and on-prem computing.

What are the advantages and disadvantages of cloud containers?

While most organisations would benefit from using portable, low-cost cloud containers, enterprises must carefully weigh up the advantages and disadvantages of a full-blown containerisation strategy.


More efficient: Without the need to boot an OS or load libraries, they can start in seconds.

Reduced overhead: Due to their sharing of the host OS, all cloud containers’ maintenance such as patching and updating can be accomplished in one event, rather than many.

Lightweight: With all OS elements virtualised, including CPU, memory, file storage and network connections, these containers take up little room on the cloud.

Portable: Because they allow IT teams to abstract application code from the underlying infrastructure, cloud containers are compatible with any platform and can run across many deployment environments.

Improved utilisation: As containers enable microservice architectures, if a single application component is struggling, just that one element can be scaled up to handle its load, rather than the entire monolithic application.



Constraints: Containers must define only one OS within each container, so they can only be run on that specific OS.

Management: When an organisation decides to containerise their IT landscape, these simple containers can quickly add up into the hundreds, which makes managing updates or patches a complicated task. It also makes visibility difficult: The sheer number of containers make it challenging to see what is going on in each container.

Security: Although cloud containers provide some protection, they are by no means unbreachable. Software inside the container may have vulnerabilities and accessing the underlying OS requires root privileges that can be compromised.

How are cloud containers used?

There are several ways to make use of cloud containers.

  • Microservices: When applications are made up of many independent services, the size and light weight of containers are perfectly suited for collecting the loosely associated services together.
  • Modernisation: Completing a digital transformation commonly begins with some form of containerisation, especially when migrating applications to the cloud.
  • Cloud-native applications: Because of containers’ low overhead/resource uptake, they can be packed quite tightly on one OS. This high density allows many of the containers to be hosted within one virtual machine, which is ideal for cloud-native application delivery.
  • Migration: Lifting and shifting applications to the cloud is much easier when they are packaged into containers because they can often be moved without changing any code.
  • Batch processing: As organisations seek greater efficiency, they implement batch processing to execute activities without human intervention, which is made much easier using containers that don’t require individual environment or dependencies management.
  • Machine learning: Data scientists can run individual algorithms within separate containers, which makes the process of machine learning efficient and easily scalable.
  • Hybrid, multicloud: When organisations operate across multiple clouds in conjunction with their own data centre, using containers makes the most sense because of their ability to run consistently across every type of environment, whether on-prem, laptop or cloud.

HPE and cloud containers

HPE, an edge-to-cloud company, offers a vertical software stack to help customers with application modernisation, virtualisation and cloud migration using various deployment models. This includes best practices on how to integrate containers into their digital landscape.

For example, HPE Ezmeral is a software portfolio that includes a solution dedicated to container orchestration and management. The HPE Ezmeral Container Platform provides the ability to deploy and manage containerised applications on any infrastructure with emphasis on bare metal (while also supporting VMs) in enterprise data centres, colocation facilities, multiple public clouds and at the edge. Customers can run cloud-native or non-cloud-native applications in containers without refactoring, manage multiple Kubernetes clusters with a unified control plane and leverage a high-performance distributed file system for persistent data and stateful applications.

In addition, HPE Ezmeral Runtime is a software platform built entirely on the open-source container orchestration platform Kubernetes that helps organisations simplify analytics, DataOps and app modernisation. It is designed for both cloud-native and non-cloud-native application deployment running on any infrastructure environment, whether on-prem or in the cloud.

HPE has also expanded the capabilities and support of the most popular open-source data analytics and engineering offering, Apache Spark. By integrating the open-source Apache Spark 3.x operator into our container platform, HPE is the first and only company to deliver enterprise, on-prem Apache Spark on K8s, “EZ Enterprise Apache Spark 3.0 on Kubernetes (K8s)”.

The HPE Ezmeral Container Platform and Runtime are available as cloud services through HPE GreenLake.