You are already using container solutions and would like to have
your systems professionally analyzed and optimized?
Or you are considering deploying scalable
container cluster infrastructures to replace
existing environments and need professional advice?
Then you are at the right place.
A few words on container clustering and orchestration:
You'll surely already have noticed the
trend reversal, which refrains from deploying complete virtual
machines (VMs) and instead relies on lightweight containers with
explicit applications.
Even though containers, and in particular dockers, were carried up
very fast on a wave that was sometimes artificially heavily hyped,
containers have become an integral part of a new IT landscape
with a new concept: the so-called microservices.
In a nutshell, the goal is to decouple server applications, deploy them as
simple and easily as possible, keep them
highly available, and scale them seamlessly during operation.
Containers are still largely based on Docker.
However, as an alternate container engine for adult orchestration
solutions such as Kubernetes, the CRI-O project (Container Runtime
Interface - OCI Standard), launched by the Kubernetes project and
supported by Red Hat, SUSE, IBM, Intel and other industry players,
is already available in
the starting blocks.
Even at the smallest level, Kubernetes offers a more efficient
strategy of deploying containers with its pod-concept
than other providers with their "standalone"
containers.
In addition, Kubernetes offers more and powerful
orchestration qualities, e. g. directly
compared to Docker (Swarm).
One more reason why Docker decided in late 2017 to offer Kubernetes as an
alternative orchestrator for its EE / DDC platforms instead of
Swarm.
Kubernetes (K8s), which originates from Google's "Borg" orchestration
platform, and now hosted by the Cloud Native
Computing Foundation (CNCF), is one of
the most powerful and widespread orchestration platforms in the
world.
In combination with Kubernetes (K8s), pods/containers
can perform many tasks: complete, self-monitoring, highly available
and scalable container clusters with intelligent key / value
backends, fully automatic service discovery, even rolling upgrades /
rollbacks during operation, unnoticed by users.
Features such as multiple release tracks, multi-tenancy
capabilities, network policies and RBAC, as well as the intelligent
operator resources, which
are able to rollout and scale clusters of
stateful applications such as databases, key/value stores, or even
SDS solutions in the K8s-cluster,
make Kubernetes one of the most powerful container orchestration
solutions on the market.
Flexible, up-to-date SDS (Software Defined Storage) solutions such
as
Ceph or Gluster can seamlessly integrated with
Kubernetes environments, to provide optimal
and cross-cluster, persistent data storage.
Storage Classes and Claims offer the ability to transparently
integrate storage tiering models.
As a Platform-as-a-Service (PaaS) solution, Red Hats OpenShift
(which is largely based on Kubernetes,
and Red Hat is an upstream contributor to
Kubernetes) offers several predefined
resources such
as
router (ingress) objects, or the possibility of S2I (Source to Image
Builds).
Red Hat's OpenShift adds functionality to manage the complete
container lifecycle and provides complete build, CI / CD (Continuous
Integration / Delivery
) and
orchestration techniques.
This creates an environment that transparently supports DevOps
principles such as faster build-to-market and continuous delivery
for dev like ops.
[click
here to get back]
|