Why open source is a key enabler for hybrid multi-cloud – but cloud must pay it back (Guest blog from OVHcloud)

Author: Rémy Vandepoel, Technical Cloud Evanglist, OVHcloud

According to the Linux Foundation, 47% of organisations in Europe have seen an increase in value from open-source technologies in the last year, and one of the major areas where open source has seen a significant shift in uptake is cloud. A recent study by O'Reilly found that 70% of enterprises prefer cloud providers based on open-source technology - and since 90% of all cloud infrastructure now runs on Linux in some way, it's clear that open source is a good fit for cloud.

However, with many organisations using different cloud providers for different workloads, open source may not necessarily seem like a natural first choice; it often requires more support to maintain and seems less straightforward, even if it’s more interoperable.

So, what’s each side really got going for it?

Open source, open possibilities

Proprietary technologies can often seem like a no-brainer. They’re from big vendors and it feels like a safe pair of hands – rather like the old saying goes, you won’t get fired for buying IBM (although in turn, it’s well worth noting IBM’s acquisition of open-source champion Red Hat in 2019!).

But many organisations find that their safe pair of hands become rather less suitable as feature creep sets in, packages become bloated and slow – not to mention that product roadmaps aren’t guaranteed to travel in the right direction for your organization. At the same time, businesses run the risk of vendor lock-in, because high discounts often come with a price: long contracts. Similarly, when it comes to cloud, if you have to pay a license fee every time you spin up a new instance – not to mention the additive hosting costs – the TCO can quickly add up. 

For these reasons, the open-source community, particularly in cloud, is still thriving. According to the Red Hat developer survey, over three-quarters (77%) of organisations are planning to increase their open source adoption in the next 12 months. Open-source software is essentially free, so there is no increase in cost per CPU or cluster. Furthermore, cloud environments like OpenStack are built for standard infrastructure, where other cloud services may require specialized technology. Clearly, in some cases, workloads do run best on specialized infrastructure, but it’s always good to know what your options are.

Open source also benefits from thriving user communities sharing best practice and tools – if you have a particular challenge with your environment, there’s a good chance that another organization or developer has a similar challenge and will be able to share a solution, code, or at least the start of an approach to solving the issue. The flip side of this is that proprietary solutions will often have support built into agreements (at a cost, of course) but open source arrives ‘as is’. This means that external support for open source can takes longer to procure – because it’s not available by default – but that doesn’t have to mean it's more expensive.

Furthermore, many open-source platforms are inherently portable. Technologies like Docker and Kubernetes were developed to provide an ‘all-in-one’ environment where the dependencies and libraries required by application components are all wrapped up and included in the container, making it easier to move, replicate or migrate functionality. This is a near-perfect fit for hybrid multi-cloud environments where having mobile application components is a huge boon.

Challenges and Responsibilities

Clearly, open source is not and has never been a silver bullet. First and foremost, it can be very difficult to ‘lift and shift’; re-engineering and moving an existing application from a proprietary environment to open-source cloud infrastructure requires an almost unfeasible amount of resource and collaboration. Furthermore, creating an open-source environment in-house can be exceptionally difficult – for example, it’s much easier to work with an OpenStack provider than trying to run and manage it internally, because of the time, human resource and investment required.

There are also pitfalls that apply to both open- and closed-source infrastructure solutions when it comes to hybrid multi-cloud. For example, it’s important not to put different application components in different clouds, because it’ll significantly slow down the functionality of the app in question.

And there’s also an element of responsibility. Open-source communities are just that: communities, whereas open-source providers, partners and support organisations usually exist to make money. This can often result in a one-way benefit, as communities put in the time and hours to create stable releases, share best practice and tools, which providers then convert to profit.

It’s crucial that vendors support the open-source movements that they are leveraging. This can come in the form of contributions, code or promotional events, but it’s vital that it’s a two-way street rather than a commercial organization profiting from a community project based on goodwill and a genuine desire for innovation.

Ultimately, there are strengths and weaknesses to both open- and closed-source cloud solutions, but in the coming age of hybrid multi-cloud, open source has a great deal to offer in terms of flexibility, financial freedom, and community support, not to mention the cloud-native open-source technologies that are intrinsically portable and agile.

However, communities need support, and it would be irresponsible of the vendor and partner community to simply take advantage of the goodwill of others. We must support open source, or at some stage in the future, we may find its doors – counter-intuitively – closed.

Authors

Rémy Vandepoel

Technical Cloud Evanglist, OVHcloud