Recent database offerings by AWS -Good for users, Dangerous for open source business models

Written by ashan.fernando | Published 2019/01/29
Tech Story Tags: cloud-computing | open-source | saas | business-models | competition

TLDRvia the TL;DR App

Amazon Web Services (AWS) is one of the leading cloud providers who offer a range of cloud services to their customers under their portfolio. Although many of the AWS cloud service offerings are built in-house, some of them are fully managed open source software solutions. In addition, few of there in-house service offerings are also compatible with well-known open source compatibility drivers, interfaces & etc.

Open Source and Managed Services in Cloud

Over the past decade, leading cloud providers including AWS has moved their focus in providing higher-level services. These include not only cloud-specific services but also popular open source software offerings.

You might be wondering whether this is something happened recently. In fact, AWS was offering fully managed Redis (open source in-memory database) from 2013 under their service portfolio categorized under Amazon Elastic Cache service. This was later followed up with Maria DB (MySQL open source variant under Amazon RDS), ElasticSearch (Amazon Elastic Search), Kafka (Amazon Managed Streaming for Kafka), Apache ActiveMQ (Amazon MQ) & etc.

In addition to the managed service offerings, AWS also came up with their native implementations (or using custom builds of open source software) which could replace existing open source software by having compatible interfaces to their open source counterparts. One of the earliest examples for this is the introduction of Amazon Aurora enterprise-scale database with MySQL and Postgres compatible drivers and connectors. If we consider Aurora database, as AWS claims, the internal implementation is totally different but the interfaces are the same providing the possibility to migrate existing applications as well as to use the already built eco-system around them. The most recent announcement of MongoDB compatible database offering named Amazon DocumentDB took lots of attention from the open source community as well as the AWS users. Amazon DocumentDB started by supporting MongoDB 3.6 API which was using open source Apache 2.0 license from its inception. This news flabbergasted most of the MongoDB users running their MongoDB cluster in AWS and managing it themselves.

Open Source Business Models

Although it was great for the cloud users to get all these open source software fully managed in AWS, it was not the same for some of the open source businesses. This was mainly due to different business models adopted by the business behind open source software.

Let’s look at two open source business models that get the most hit for their business when AWS or any other cloud provider takes advantage of their open source licensing models.

  • Premium or Production Support — Offering open source software free of charge and having production support plans.
  • Managed Service Hosting — Providing hosted and fully managed versions of the open source software in the cloud for a premium price.

Premium or Production Support

If we consider the business model with production support, very large cloud providers such as AWS is fully capable of forming their own specialized team within AWS. This has a direct impact on customer support business models since the team within AWS could also integrate with the AWS cloud support plans and provide a short feedback loop.

Managed Service Hosting

If we consider AWS offering fully managed versions of the open source software as a service, this could have a direct impact on open source companies who are already doing this as their core business. This was the case that happened to MongoDB very recently. MongoDB was offering fully managed MongoDB service offerings with premium support and once AWS offered their own fully managed version of MongoDB, this has introduced direct competition.

Open Source Business Challenged by Cloud

If we looked at the MongoDB incident, they have re-licensed their open source database under a custom license which explicitly limits the ability to provide MongoDB as Software as a Service (SaaS) offerings. This is not healthy for open source movement since it limits the freedom of usage but the situation was expected from MongoDB side where they have faced a dip in market value as soon as AWS announced this service offering.

If we look at history, this was not the first time re-licensing happened as a result of similar incidents. Redis also did similarly in the past when other cloud providers started offering Redis as managed services.

Unfortunately, we are right now in a situation where we have an increasing trend of happening this again. This is because the cloud providers are constantly pushed towards offering managed services by their users while open source companies get affected by their business models. Looking back at the incidents happened with AWS, it is quite unsure whether they will keep the partnerships or move to their own offerings.

Solving the Problem

However, the direct way out is to build sustainable business models between Cloud providers and Open Source businesses. This could be lead by building strong partnerships with open source businesses when offering their software as managed services in the cloud. The staggering fact is that some of these open source businesses were partners with AWS just before AWS decided to go on their own.

When considering the MongoDB incident, many others claim it is more like a stab in the back. However, without knowing the details behind what has happened, whether there were any attempts of negotiations from AWS with MongoDB, it is unfair to point fingers at any of them.

But I strongly hope that this trend of conflict will end in future allowing to build healthy relationships with open source businesses and cloud providers.


Published by HackerNoon on 2019/01/29