Kubernetes Cloud Hosting For Better App Management

HC Blog

There are various methods to deliver application services with high uptime. VMware Could computing technology is one of the best options to create, provision, and manage cloud servers with High Availability or Fault-Tolerance. However, application delivery isn't about infrastructure uptime. It is about application uptime. So, even if the infrastructure technology layer of certain software applications is up and running the app can still be down for various reasons. It is could be a load on the web server or the database for example. So, besides the infrastructure reliability and uptime, it is very important to deliver an application hosting environment with the highest possible uptime. Here comes Kubernetes.

Kubernetes Cloud Hosting Environment

Let's start first with a short explanation of what Kubernetes is. According to the Kubernetes themselves, it is a "software platform for managing containerized workloads and services, that facilitates both declarative configuration and automation". Google has started the Kubernetes project in 2014.

The traditional application hosting environment runs either on top of bare-metal dedicated servers or on virtualized (including a Cloud-based) environment. When deployed on bare-metal dedicated servers there is only a certain operating system between the software apps and the hardware. This is a traditional and proven method of software delivery. It is not scalable, however. As soon as the installed software reaches the resource limit of the underlying physical server, the application delivery stops.

Cloud-based (virtual) environments offer much more salability and flexibility. There are different types of virtualization technologies. When full virtualization is used (VMware or Kernel-based Virtual Machine for example), the software applications run in an OS inside a virtual server, which themselves are created through a hypervisor on bare-metal servers. The virtualized environment is more complex. There is an operating system (OS) for each Virtual Machine that creates a certain application environment. There is also an underlying OS that powers the virtualization (or Cloud) environment itself. Such an architecture allows an abstraction of computing resources. Applications can be deployed in different VMs, scaled, migrated easily.

Kubernetes, the main topic of this publication represents another approach to creating and managing an application environment. It is a form of virtualization (abstraction of computing resources) known as OS virtualization. The OS virtualization itself is done on an operating system level, on top of the underlying bare-metal server. Instead of using hypervisor (full virtualization) to create virtual machines, OS virtualization (OpenVZ, LXC, Docker, etc.) uses the kernel to make computing instances called containers.

Containers share the resources of a single operating system (OS). Therefore the applications installed on them also share the computing resources. Each container has its own filesystem, a certain amount of allocated CPU and memory, day storage quota, and more. OS virtualization has its advantages. Although it is definitely not a better approach to cloud computing compared to the full virtualization, used with Kubernetes as an orchestration technology, containers offer certain advantages. Kubernetes itself can be run also on Virtual Machines, so the underlying virtualization technology is a preference.

What is important about Kubernetes is that it applies an application-centric approach to cloud computing infrastructures. It creates a distributed infrastructure environment (for distributed web hosting read the article "Distributed Web Hosting for Web 3.0 Websites"). Kubernetes arranges the deployment, management, and scaling of software applications. On an infrastructure level, it allows for creating application containers across a cluster of either physical or virtual servers. Kubernetes offers:

Application-level load balancing: Kubernetes makes containers visible through a DNS name or IP address. If the load on a certain container is high, it distributes the queries and traffic to other containers.
Storage planning and coordination: Storage systems can be mounted into Kubernetes-based architectures.
Automated System Scalability: Kubernetes allows for rollouts and rollbacks of the IT infrastructure based on the resources demand. Certain system conditions can be set. Based on it Kubernetes automates the infrastructure environment by creating new or removing existing containers from the infrastructure.
Resource utilization: Base CPU and memory limits can be set for each container in a Kubernetes cluster. It runs containerized tasks and application delivery with the best possible resource utilization.
Failover and application continuity: Kubernetes has a failover function that reboots and returns online any failed containers. It can also replace them and take down containers that don't comply with the rules predefined by administrators.
Security and reliability: Through Kubernetes, administrators can store and maintain sensitive information - passwords, OAuth tokens, SSH keys, etc. Application configurations can be set and updated without a need to rebuild container images and without exposing sensitive data.

What HostColor does with Kubernetes?

We use Kubernetes to provide a unified environment for application development and application delivery. We deploy and provision Kubernetes-based Public and Private IT infrastructures - either based on bare-metal server or your preferred enterprise virtualization (Cloud) technology. We run Kubernetes clusters on VMware ESXi, Proxmox, Kernel-based Virtual Machine, and Docker.

Kubernetes Private or Public Clouds are deployed for HC customers in 40 locations in North America, Europe, and Asia.