The value of containers has been recognised for millennia. From sealed earthenware jars used to preserve food in ancient Egypt, to the millions of ISO shipping containers travelling across sea and rail routes that have fed the globalised world of today. In the modern IT era, containerisation offers a form of operating system virtualisation that allows developers to package up all or part of an application, effectively into a portable building block.

 

The use of containers has grown rapidly over the last decade with analyst firm The 451 estimating that around half of all enterprises now use the technology. Docker has emerged as the most popular container form; however, the open source pioneer is only half of the story. For containers to really shine they need additional platforms to streamline scheduling deployment, storage management and scaling which is where Kubernetes has emerged as the most valued orchestration tool.

 

Together, the two standards allow developers to create applications that are lightweight, portable and highly efficient. Big users like Google literally create billions of them each day to serve Search, Gmail, Google Docs and a whole host of other cloud applications. And when you don’t need them, they just evaporate, releasing resources for the next user.

Perfect for functions and services

One of the big advantages of containers is that they are incredibly efficient. However, they are by design, stateless in that they don’t store data after they have been torn down – this makes them perfect for single functions and services.

 

However, data storage is still the main inhibitor to the wider use of containers as there are few truly stateless applications within the enterprise. For more complex containerised applications that maintain databases, this statelessness can lead to issues such as limited disaster recovery capabilities or poor performance in situations where containers may need to restart frequently.

 

Another challenge is cloud portability. Although containers by their very nature can be used across different environments, the big cloud vendors namely Amazon, Google and Azure have slightly different implementations of the orchestration tools. In addition, how containers interact with storage elements is not uniform.

 

Instead of debating about whether to use stateless or stateful containers, or which orchestration platform is better; the most beneficial approach is to do both with every cloud – to have your cake and eat it too! This is exactly the philosophy that NetApp is pioneering with Trident and the NetApp Kubernetes Service as part of their Data Fabric approach.

 

In simple terms, the solution acts as a storage provisioner and orchestrator that takes advantage of the StorageClass API object in Kubernetes to consume storage resources on demand, without having to worry about the details.

Deployable in three clicks

As a true cloud-native service accessed through a unified portal, an entire Kubernetes cluster can literally be deployed in three clicks without the need to download any software. The cluster can utilise both modern object-based storage and legacy NFS and SMB storage pools and includes useful capabilities such as backups, clones and synchronisation that are accessed from a single management console.

 

The solution uses the open source version of Kubernetes and Istio which makes it easy to create a network of deployed services with load balancing, service-to-service authentication, monitoring, and more, without any changes in service code.

What’s the catch?

But the real kicker is that it is natively multicloud. This allows developers to spin up containerised clusters across AWS, Azure, Google Cloud and a whole host of other vendors, plus support for bare metal with the ability to move clusters between provider without technical limitations. Cynically-minded types might be asking, “Yeah, great, but what’s the catch?”. The simple answer is none and as a testament to that statement, we are delivering it as a simple pay-as-you-go SaaS model. To allow developers to test drive it for themselves, there is a 30-day free trial (no credit card required).

 

Although only recently introduced to the NetApp portfolio, developers using the service’s capabilities are already reporting up to a 90% reduction in the time it takes to deploy and maintain clusters – especially for time-consuming tasks like patching.

 

According to the analysts, the median number of containers per host rose by 50% over the last year, and trajectory looks similar for the next few years. By recognising that freedom and flexibility is vital for the next generation of cloud services to flourish, a true multicloud approach that simplifies deployment and ongoing management will be the cherry on top.

 

Why not try the NetApp Kubernetes Service for yourself today with a 30-day free trial (no credit card required).

Martin Warren

Martin Warren is EMEA Cloud Solutions Manager at NetApp. Based in the UK, Martin has many years of experience working in data protection, data storage, virtualization, cloud, big data and networks. In his current role, Martin is responsible for NetApp’s EMEA cloud strategy and solutions with a focus on driving business growth and aligning NetApp’s cloud offering with customer and partner demand.

Martin is a specialist in private, public and hybrid cloud solutions, including the aspects of data storage and data management. He provides advice to businesses on the advantages and impact of different cloud and data storage solutions, helping them to meet their goals. In addition, Martin is a subject matter expert on the EU’s General Data Protection Regulation (GDPR).

Prior to joining NetApp, Martin held positions at Symantec, Sun and StorageTek. He also worked at IT service delivery and consultancy companies Misys and 4Front Services, advising businesses on adopting IT services specifically from a blended approach.