olly - Fotolia

What the AWS and open source spat means for users and enterprises

Explore the concerns at the center of the AWS vs. open source debate and how they play into a larger conflict between the open source community and public cloud providers.

When AWS rolled out DocumentDB and its Open Distro for Elasticsearch, the open source community pushed back. Indeed, MongoDB and Elastic themselves seemed to lead the charge.

Many enterprises depend upon AWS and open source software (OSS) -- either on premises or in the cloud. And while technology fights are always interesting to watch, users need to know where the software they rely on stands. They want a pragmatic set of guidelines and advice about what this means to them.

What is at stake for AWS and open source

There have been questions about AWS' commitment to the open source community for years, and the issue once again made headlines when Amazon added DocumentDB, a MongoDB-compatible database service. Things came to a head in March with Open Distro for Elasticsearch, which takes AWS' recent addition to ElasticSearch and provides it under the open source Apache 2.0 license.

The concern with Open Distro is that AWS is forking and redistributing an open source version of Elasticsearch, customized for use within AWS and for AWS users.

If you believe in the spirit of open source, you might think this move by AWS seems fishy. AWS disagrees, of course. Indeed, it claims that Open Distro for Elasticsearch only uses the open source code for Elasticsearch and Kibana. AWS stated that its "intention is not to fork" and will continue to send contributions and patches to the common code tree. While we have to take AWS' word for the fork-or-no-fork controversy, both options are a common practice for commercial software companies.

Enterprises that use AWS and on-premises open source systems might need to get accustomed to this situation.

Open source remains somewhat in conflict with public cloud providers. Through their managed services, cloud providers offer much easier and faster ways to consume OSS than licensed open source systems from MongoDB, Elastic and others. Plus, many of these OSS vendors offer their own managed versions of their software that are available to host on the major public cloud platforms.

Expect the number of OSS-based products to increase as cloud users demand them as platform analogs for applications they move to the cloud. In that sense, cloud providers offer a valuable service when you consider they remove the need to own and maintain hardware. AWS and other cloud providers also take OSS in new directions that tend to overshadow the past efforts of the open source contributors.

AWS vs. open source: The cloud provider perspective

Those in the open source community view Open Distro as light exploitation, or in the most favorable light, they think it muddies the open source waters. Of course, the public cloud providers see it differently for two key reasons.

First, AWS needs the ability to make basic open source systems multi-tenant. Most open source systems are not built for public cloud, where multi-tenant and multiuser functions must exist. Thus, public cloud providers must optimize or extend some parts of the open source system to provide cloud-native features, such as memory, CPU and storage management. And since these features are unique to each public cloud, they won't add much value when you put them back into the code tree.

Second -- and perhaps what's causing the dust-up here -- is that public cloud providers need to differentiate their version of the open source distribution to earn a unique foothold in the market. For instance, an open source system may come with 50 core features that define its value within the marketplace. However, users who need that open source system in the public cloud demand cloud-native integration with their provider's security, monitoring and management, storage and databases.

The tighter the cloud providers integrate these things with hosted OSS, the more they add code that is valueless to the open source community. This participation with little contribution back in terms of open source value is what drives the AWS vs. open source debate.

Measured enterprise concern

For enterprises that use AWS and OSS, there are several ways to look at this. If an enterprise uses a managed service, such as DocumentDB, it can mostly abstract itself away from the open source politics.

Those that use open source systems on premises or even within other public cloud providers may have the most to lose. If a large public cloud provider takes the system in a new direction with its own extensions or introduces a true fork, then those that rely on the pure open source system could end up holding onto an outdated and less developed open source code tree.

Still, even those users shouldn't get too caught up with this latest technology dust-up. These conflicts are typically short-lived, and the impacts end up being minimal.

Dig Deeper on AWS cloud development