Covid-19 has had a significant impact on most aspects of our lives in 2020. Although some companies saw unexpected growth due to offering in-demand tools or services, I think most of us will be happy to draw a line under 2020 and hope for a better 2021.
In the spirit of 2021 optimism, I have been considering the current state of the open source database market and putting together my Big Bold Open Source Predictions for the next 12 months.
Big Bold Prediction One:
I believe reliance on classic operational roles such as sysadmins, database administrators (DBAs), and web admins will continue to shrink with the drive towards “easy hosting” and “as a service” models for running databases.
We will continue to see the sysadmin, DBA, and web admin roles evolve into site reliability engineers (SREs) and specialists instead. We will also see architects, developers, and non-infrastructure experts continue to pick the backend stacks. This furthers what I call, the technology inheritance problem, where you have to run and support legacy tech stacks. While they might have originally been picked for good reasons, those reasons might not still apply today.
As these changes accelerate, the demand for specialized expertise on how to build, design, and architect databases will grow. The “application” database expert will be an in-demand role in the future. This will force businesses to fill the gap with consultants or on-staff expertise, or risk increased costs and lock-in over time.
Big Bold Prediction Two:
Open source software has ‘won!’ Most new software projects will include at least some open source software components, and the majority of new releases will take place under open source licenses. However, at the same time, there is a reduction in the level of openness taking place. Classic software infrastructure vendors will accelerate putting additional controls in place around how their open source projects can be used, but cloud providers will race to embrace a more open model.
It’s weird that companies like Amazon are now releasing more open source, while traditional open source vendors like MongoDB, Elastic, Redis, and others are introducing more restrictive licenses and “as a service” offerings that actively lock people in. Many database companies’ newest and most innovative features are starting to be reserved as “DBaaS Exclusives”.
Ironically the push to retain market share by so-called “open source” vendors will actually drive people away from their platforms. Look back at Oracle in the ’90s and ’00s… Their introduction of price hikes, lock-ins, and other restrictive and expensive methods, forced people to look for alternatives. The good news here is that we may be on the verge of a new wave of disruptive innovation to deal with this trend.
Non-open “open source” sounds like an oxymoron, but it is really becoming a thing! At Percona we believe in true open source, not in versions that are crippled by design and act as a gateway drug to get you hooked, unable to escape, and filled with regret.
Big Bold Prediction Three:
Consumers and users will revolt against these open source shenanigans. Docker and Red Hat are good examples here – they have taken open source projects or products and then limited them in order to get more people to upgrade to the “Enterprise” version. This then in turn leads to potential lock-in.
Docker started playing games, locking people out of downloading images on Docker hub, and trying to generate a business model that forced people/vendors to pay for access. Red Hat has changed its approach to CentOS from a fixed release model with long term support, through to a streamed model with no such certainty.
Where will this lead in 2021? People will find alternatives. Kubernetes is depreciating support for Docker and vendors are stepping up with alternative catalogs. A fork of CentOS is already available with Rocky Linux.
Big Bold Prediction Four:
Databases will continue to move towards a run-anywhere-on-anything reality. Users want DBaaS type solutions, but want to avoid sole vendor lock-in. They are considering technologies like Kubernetes to manage this.
Enterprise companies (Nutanix for instance) are looking to bridge database deployments across cloud providers and data centers. Open source companies like Red Hat are pushing the option of running your databases like you are already running your application, in a cloud-native way.
Of course, all cloud providers are now digging into the multiverse, with Google Anthos, Azure Arc, and AWS EKS available anywhere. At AWS re:Invent 2020, they announced the decision to allow companies to sell professional services via AWS Marketplace. This signals that even AWS knows that users committing to a single provider is unrealistic and that users will still want advice, consulting, and support from other vendors.
What makes this even more interesting is that database providers can offer “single database” and “any hosting provider” as a service offering. This means that more friction will materialize between the two groups.
Eventually, some of the cloud providers will use their market cap and purchase some of the open-source vendors in the space. MongoDB as a Microsoft company? Elastic as an AWS company? Yep! It will happen, maybe not in 2021, but this consolidation is bound to occur as it becomes too tempting to acquire resources to create value.
Alongside all this, getting true open source DBaaS to market will be important. Users want the flexibility of running their databases in the cloud, but they will also want an easy route out of using a particular cloud service if they choose to move. For cloud providers, facilitating this open approach will allow them to meet the needs of their customers, and ensure the relationship is based on mutual respect.
Big Bold Prediction Five:
Increasing innovation will come from the data management space. There is simply too much data, too many applications, and too much sprawl to not necessitate innovation in this space.
We have seen rapid growth in the number of technologies developed to deal with the increase in databases and data technology. For instance, the growth in the troubleshooting and observability space over the last five years has been unprecedented. Over the next twelve months, this will continue to develop rapidly.
No company ever says; “I want to have less data.” Instead, technology teams will need to ensure they can get value out of their data more efficiently, to justify the cost of managing all that data over time. Optimizing this spend, and how data is stored, will go hand-in-hand.
So, these are my Big Bold Open Source Predictions for 2021! Do you think these are on track, or do you think the open source database market is going to move in another direction?
