Data Backup Digest

A container is an application rolled up into a single package. It contains all the dependencies, libraries, binaries and configuration files that are necessary for the application to run and function correctly. Because it’s a single package, it means that it can easily be moved from one environment to another without things breaking or going missing.

The environments it moves between can be many different things. For example, it could be when moving from testing to live or from a physical machine to a virtual one. Containers help navigate the tricky territory of different operating systems, versions, protocols, and more.

Containers are different from virtualisation, though both are born from the same theories. While virtualisation is replication of an entire machine and its operating system, a container is smaller and shares the operating system with other containers. They don’t need virtual memory or system resources, thus consume less power and boot faster.

The fact that data containers are so light, flexible and portable means that many have begun using them in cloud applications. As storage, and computing in general, move into the cloud, containers are going to increase in popularity.

Data held within containers has no state. That means it will remain identical, no matter where it’s moved to or how many times it’s moved. That’s great for testing or analysing, but also efficient when it comes to spinning up new environments.

Big companies like Google and Spotify make use of data containers. The latter used it for container management platform Docker, achieving reduced CPU overheads. Google apparently creates around two billion containers every week.

Using storage containers can help you reduce costs. Dynamic delivery simplifies operations and reduces the need to over allocate into storage pools. Because of that it reduces the need for persistent storage and don’t need volumes associated. Redundant data can be shared across the containers, which spreads the risk. They also require less storage capacity overall, which is always a great thing – using immutable container images reduces this and is especially great for backup. This is because setting aside storage for images of the operating system is no longer necessary.

"Containers and Kubernetes improve operational efficiency, not just on the server side, but on the storage side by simplifying deployment and enabling applications to move between VMs and physical servers, which could make I/O profiles more storage friendly," said Stanley Zaffos, senior vice president of product marketing at Infinidat. The pairing also helps present a more competitive storage market, as pressure is put on vendors to reduce their pricing in an ever-growing and crowded market.

Containers are unlikely to replace virtualisation, at least not in the short time. They have their downsides. Though sharing an operating system is a positive, on the other hand it also means they are more open to vulnerabilities. Data could be leaked through the operating system’s security flaws or malicious code could be purposely injected into it. Nevertheless, deploying virtual machines and containers together, rather than favouring one over the other, is a good solution.