The idea of companies being ‘born in the cloud’ has become popularised in recent years.
Cloud computing technology has existed long enough that not only do some of the most exciting start-ups hold this badge of honour, but well-established leaders.
Australian buy now, pay later (BNPL) success story Zip Co. is a great example of this – the company has highlighted its ‘born-in-the-cloud’ status, relying on public cloud to provide computing power and storage and management of key customer and other business data. In this model, there are no hardware constraints that could become a barrier to scalability – just elastic cloud to grow in Australia’s one-to-watch industry following the record $39 billion acquisition of fellow BNPL player Afterpay.
This new generation of companies is following the idea of cloud native, with a direct and increasing impact on how the very applications we rely on to live, work, and play are being built.
Cloud native applications are described by the Cloud Native Computing Foundation as ‘scalable apps in modern, dynamic environments, such as public, private, and hybrid clouds’. They are built to be decoupled from limiting constraints and can suit the needs of companies who want to start and expand in the cloud as well as others transitioning to it. New technologies such as containerisation, Kubernetes, serverless computing, and microservices are key drivers of this shift.
But the shift can’t be confined just to the apps themselves. Beneath the service of all apps we use is an infrastructure layer that supports them. Just as you may not see anything but the sink, toilet, and shower in a bathroom, you know that none of it would work without the pipes underneath.
Infrastructure must also become cloud native. Old infrastructure was built for the old way of building applications, and even the old way companies grew. It does not fit the new digital unicorn-filled Australia we’re seeing unfold in front of us.
Cloud native storage carries similar traits to the apps it supports – it must be portable, scalable, and dynamic. It must be equipped to properly manage data for stateful apps: i.e., a program that saves data from one piece of activity for seamless use in another, while also addressing data storage challenges that can occur in the promised freedom to scale of cloud native environments.
There are six key operational areas in which cloud native storage must rank highly to match up to the cloud native world.
Companies in hypergrowth can’t sit around waiting for long sales cycles to end to procure new infrastructure that can take months to deploy. Cloud native applications can be built quickly to overcome this challenge, but you also need cloud native storage that can easily scale to manage the data that comes from these fast-growing apps and companies. More technically, scalability can be broken into four key areas: client scalability (number of users accessing the system), capacity scalability (growing storage capacity in a deployment), throughput scalability (increasing the amount of data processed per second), and cluster availability (deploying additional components to a storage cluster).
Unsurprisingly, performance matters in a cloud native environment. Essentially, systems need to work well and performance needs to be predictable despite the lack of predictability in growth and app demand. This means being able to rapidly complete read or write operations, execute a high number of storage operations per second, and deliver strong throughput rate for fast storage or retrieval.
Consistency is key. This can be defined in an IT environment by whether read operations quickly return the correct data after it’s written, updated, or deleted. If data is available immediately after it’s been changed, it’s extremely consistent. Any lag moves the dial to only eventually consistent. This makes read delay a recovery point objective (RPO).
The damage from data loss has become all too real for Australia, with major broadcasters, government departments, and other high-profile organisations all falling victim to breaches in recent months. A report from Cloudian found that many traditional defences are failing to secure data from the scourge of ransomware, prompting the need for an immutable backup copy of data. Cloud native storage must be durable against the increasing threat of cybercrime. This is about more than just access – durable platforms should guarantee that data can be safely stored for extended periods of time. That requires multiple layers of data protection, multiple levels of redundancy, and the ability to identify data corruption and automatically restore or reconstruct data.
As highlighted above, it’s important that cloud native storage isn’t bound by the restraints of storage past. Cloud native apps are portable by definition, and to support that agility, storage systems must be deployable on demand. This means a software-defined, scale-out approach, enabling organisations to quickly grow capacity without the need for new appliances. Think of how easy it is to add a second piece of luggage to a flight (for those who can currently fly of course) – that’s about the level of simplicity needed.
Finally, cloud native storage must be highly available – data access needs to be immediate no matter where a failure has occurred. Cloud native storage systems should be able to heal and restore any failed component, manage redundant data copies on a different device, and handle failover between devices.
Companies are now built off the apps they provide and use to sustain and grow. The environment in which that happens is changing, and it’s important that the storage – the pipes – behind it all change too to suit this new reality.
Whether it’s a company transitioning, or one born in the cloud that knows nothing about a non-cloud world – and doesn’t need to – having a modern, cloud native storage foundation is essential.