DC/OS 1.12 is here: Accelerate efficiency levels to the maximum

Staying competitive nowadays is an endlessly tiresome task. The demands you have to face change constantly and you have to be agile enough to be able to deal with all of them!

The latest DC/OS release promises to do just that.

Mesosphere DC/OS, “an enhanced cloud platform for enterprises to run traditional, cloud-native, and data-driven applications on multi-cloud and edge infrastructures”, is here with a new release [1.12] and it brings some exciting tools for your everyday struggle with efficiency!

Let’s take a closer look at the new features and changes in DC/OS 1.12.

Mesosphere Kubernetes Engine

Kubernetes has emerged as the defacto standard in enabling IT to build out digital architectures to deliver on their business goals, and it is now a “must have” for whatever your organization’s level of digital maturity is.

However, the constant expansion of the Kubernetes ecosystem may contain some risks as well. According to Edward Hsu, this expansion results in “increased infrastructure and operational costs, increased security risk, and dependency on specialized hard to find talent.”

SEE ALSO: How Kubernetes transforms your business

One of the newest additions to the CD/OS platform is the Mesosphere Kubernetes Engine (MKE) which “allows IT to centralize scattered and sprawling Kubernetes clusters on a bare-metal, virtualized, or public cloud infrastructure, so you can reduce operational overhead, tighten security controls, and dramatically cut infrastructure costs.”

Most specifically, MKE features:

• Consolidated multi-Kubernetes management – A centralized self-service control plane for IT organizations to manage multiple Kubernetes clusters, on multi-cloud, datacenter, and edge infrastructures
• High-density multi-cluster pooling without virtualization – Multiple Kubernetes components, including different clusters, can fit in a single server, VM, or cloud instance allowing for 2x  or more reduction in infrastructure.
• Kubernetes lifecycle automation – Automagically applied best practices for Kubernetes installation, upgrade, high availability, performance and security.

Accelerating machine learning

Clive Humby coined the phrase “Data is the new oil” and if data is the oil, then machine learning describes the process that makes this data useful.

The waves of automation and data-driven decision making have recently started crushing on the shores of many industries as they slowly but surely make headway with their digital transformation initiatives.

There are, nonetheless, challenges in delivering machine learning models.  Edward Hsu mentioned that challenges like the requirement of specialized tools, as wells as the tremendous computational resources needed for processing massive data sets, “slow down progress and increases the risk of data leaks. That’s not to even mention highly-paid data scientists that are inefficiently utilized.”‘

SEE ALSO: Does your company have a machine learning strategy?

To help you deal with those challenges, DC/OS 1.12 features the beta availability of the Mesosphere Jupyter Service (MJS) which gives data scientists on-demand access to the popular Jupyter interactive computing environment.

According to the official blog post, “MJS pre-integrates all the tools data scientists need to be productive and lets them work directly against enterprise data sources on any infrastructure. Data scientists can use high-performance infrastructure for model development, securely share research, and confidently push models to production.”

The features include:

• Jupyter Notebooks-as-a-Service – Instead of waiting for tools and infrastructure, data scientists can hit the ground running on day one with a fully functional, interactive Data Science development environment (Jupyter Notebooks), delivered as-a-Service, pre-configured with all the tools, frameworks and libraries required
• Secure Collaboration – Instead of downloading data to their laptops for analysis, data scientists work directly on shared infrastructure, this means IT can leverage integrated security controls permitting fine-grained access to notebooks and data sources
• Machine Learning Acceleration – DC/OS replaces isolated gateway nodes typically used for model training with a general purpose resource pool, which means notebooks are integrated with high-performance cluster-computing resources and distributed analytics engines to accelerate model training

Getting started

If you are eager to try out the Mesosphere Kubernetes Engine, Mesosphere Jupyter Service (Beta), or DC/OS 1.12 dcos.io, you can sign up for a demo here.

The post DC/OS 1.12 is here: Accelerate efficiency levels to the maximum appeared first on JAXenter.