Blog details

2021 Alternatives to Elastic’s SSPL Licensing changes – How does this affect you?

Over the years working as an Elasticsearch Subject Matter Expert with Weblink Technologies, it’s not uncommon to consult with clients concerning alternative solutions in the open-source distributed search-engine space. After discussing things like ROI, Elasticsearch based product strategy, long term strategy, and total cost of ownership, some of the most common questions are: 1) “What about the AWS Hosted Elastic service” and 2) Is AWS Hosted Elastic Service an option.

In this article, I want to discuss my thoughts and recommendations on the following:
  • Who is impacted by this licensing change?
  • How you may be impacted and what are the effects?
  • What options do you have if you are currently using AWS Hosted Elastic Service
Who is impacted by this change?
Segment 1: – Hosted Cloud Service Providers that Monetize Elasticsearch as a Service

With the recent change in licensing for Elastic’s main suite of products, this will no doubt affect cloud hosting providers that offer Elasticsearch as a hosted service to its customers without a redistribution license with Elastic. The good news is Elastic has not shut the door for everyone, I will cover this later in the article. Before we get into the details of Elastic’s move to SSPL licensing, I want to take a moment to mention who I think will feel the impact of this change directly, mainly… AWS Hosted Elasticsearch service. Elastic and AWS have been in dispute for some time concerning the launch of AWS Hosted Services based on a version of Elasticsearch called the “Open Distro” which is an alternative to the commercial version of Elasticsearch. Not only have AWS been successful with its launch of this version, they have also added some level of innovation that takes advantage of some of AWS native cloud offerings like S3.

Segment 2: Hosted or Non-hosted Products with Elastic Embedded as its Core

There is also another segment of users that are affected, namely the custom product development customers. This may fall under search based, data ingestion or data analytics-based products, that ship with some version of modified Elasticsearch embedded as its core functionality. If your organization has products which uses a version of Elastic at 6.8 -7.10 or Open Distro or free elastic, it is now at risk of being forced to release its intellectual property under terms dictated by the SSPL Licensing Agreement. It is also important to understand what previous Elastic versions to which the new SSPL changes will apply. I was able to find this on the FAQ section associated with Elastic’s announcement.

“What versions does this change apply to?
This change only affects the source code — our releases will continue to be free and open under the Elastic License. This change will apply to all maintained branches of our software — 6.8, 7.x and master/8.0, and will take place before the 7.11 release is generally available.”

It’s important to fully understand the term “free and open”. Elastic has recently moved more and more of its key features under the “Basic Elastic License”. This allows end users to take advantage of many of the new features of Elastic without opting to purchase a Gold licensing package. The Basic Elastic License also protects its core features from being used to create other products based on Elastic’s original core source and software.

Why is this happening? Beating them at their own game!

Over the last few years AWS has been making some bold moves to monetize Elastic’s main offerings and basic license features. AWS recently added their own version of professional services to support their existing customers already on AWS hosted version of Elasticsearch. This was clearly a move intended to go directly after Elastic’s customer base. AWS hosted Elastic Services comes with Ranking LTR modules, SQL query interfaces, tiered storage, and advance fail over features with many easy-to-use Elastic cluster management features. We are seeing many of our clients opt to use AWS Elastic service to get their project going quickly. They are not interested in talking to the representatives at Elastic when it comes to choosing a search engine or discussing their use-cases. These clients just want to stay within the cloud platform they know best and not worry about learning Elastic’s products, licensing, or services. It is also important to point out how AWS has innovated their Elastic offerings in this space. The addition of Ultrawarm technology allowed users on AWS to reduce the cost associated with keeping large amounts of data active and ready for search. As a response, Elastic recently added a similar feature in the Elastic 7.x upgraded version releases. With innovations like this, it is easy to see why Elastic is now reevaluating how it is going to approach the future in this space.

My Organization runs AWS Elastic service- How does this affect you?

If your organization or product falls into one of the two impacted groups, then you may be inadvertently exposed to legal and licensing violations. Also, you may have to incur additional expenses associated with remedy options to address this change. Your core operation is now faced with a dilemma, either migrate to Elastic hosted services (depending on your use-cases, this can be a very non trivial task) or evaluate if managing your own cluster is the right approach. Both of these options are considered low-hanging fruit, so I thought I would mention them upfront. If your product is using Elastic underneath the covers, please contact us at Weblink. We specialize in coming up with innovative alternative solutions to reduce the overall impact of this licensing change.

To summarize the effects are as follows:
  • Your organization may need to migrate to alternative solutions (Elastic Cloud – additional considerations are needed to determine if this is the right approach given your specific use-case).
  • Your Core product will require a new strategy if Elastic is embedded and modified.
  • Potential customers that use your product may be impacted.
  • AWS may decide to discontinue the service altogether or increase the pricing directly to Elastic for its hosted service running on AWS platform.
With that being said, let’s start looking at some of your options.

What options do you have if you are currently using AWS Hosted Elastic Service?

As I stated earlier, Elastic has not shut down all the cloud providers. In fact, Elastic gives a list of current cloud providers that are authorized to provide hosted services using their commercial software which includes Microsoft, Google, Alibaba, and Tencent.

What is the best way to mitigate the impact of this change? To answer that question, I’m recommending our Elastic Product Strategy offering where we professional consulting time working with your IT team to find the right solution given your unique use-case.

Recommendation from a SME!
I am going to cover the more common options and then we can address the more complex and specialized options in a strategic consultation service call. First, I would start to explore the magnitude of how this change affects your current operations. If a migration is needed (which is most likely the case), then I would cover the following options below:
Options:
  1. Determine the Magnitude and determine if you should host your own.
  2. Migration to the Elastic’s Cloud offering.  We would still need to see how AWS responds to SSPL changes; it is too soon to tell at the time of this writing. However, in many cases, moving to Elastic’s cloud offering may be the right option, but it’s not an automatic option. It is important to ask the right questions and get a full assessment from our professional services to fully understand the impact of switching and what other options may be available. However, that being said, I would say, for very small and smaller clusters this is a viable option, no need to overthink it.
Additional topics covered during our (Elastic Product Strategy Offering)
  1. Alternative solutions to Elasticsearch – This has been increasingly requested as an option, we help to re-imagine what additional options you have for processing, searching, and perform analytics on your data sets.
  2. Embedded Elastic for OEM and core products.
  3. Complex data ingestion pipelines.
  4. Search as a Service or Search API’s
Additional Considerations:
It is important to point out, there are many topics to consider with the options mentioned above. I am not covering any topics associated with very large datasets and very large Elasticsearch clusters. In fact, my recommendations would most likely change if you run larger clusters of 25 – 200+ nodes of elastic or 1TB of data an above. For smaller implementations, these recommendations should get you started.

If you need any help with planning or strategy concerning how best to pivot, please feel free to contact us. I am more than happy to provide our services to help you achieve optimal results.