OpenSearch has been a buzz in DevOps over the first half of 2021. The project is moving forward, but understandably there are a lot of questions. This article will address some of those frequently asked questions, and will be updated to address more over time.
What is the OpenSearch Project?
OpenSearch is an open source software project launched in 2021 as a fork of the Elasticsearch and Kibana projects, with development led by Amazon Web Services. The project includes a database (also named OpenSearch) and a frontend visualization and analytics called OpenSearch Dashboards.
Why was the OpenSearch project created?
In January 2021, Elastic, the company behind the Elastic Stack (made up of the Elasticsearch, Kibana, Beats, and Logstash projects, and often known as the ELK Stack or Elastic Stack) – announced that it would shift to a dual licensing structure based on the Server Side Public License (SSPL) and the Elastic License – neither of which has been recognized as an open source license by the Open Source Initiative (OSI).
This move to a proprietary dual license created a clear demand for an open source alternative, laying the foundation for introducing OpenSearch.
Prior to the revised licensing move by Elastic, this was the case for Elasticsearch and Kibana, and the resulting incentive to utilize this code base drove huge numbers of talented developers and outstanding organizations to contribute to these projects.
With that avenue walled off, people clearly needed a new direction, one that would not be prone to these risks of vendor-owned open source, but rather a community governed OSS.
Is OpenSearch actually ‘Open Source’?
Indeed it is! The entire purpose of OpenSearch is to enable a traditional, open, and community-based approach wherein participants will maintain, safeguard, and innovate the project, which is being made available under the venerable Apache 2.0 License.
This community-driven, open source approach is being undertaken with the specific goal of driving greater innovation for future iterations of Elasticsearch and Kibana – allowing developers, enthusiasts, and creators, along with the organizations they work with, to build the best software that they can dream up!
So what’s the difference, at this point?
Differences between the ELK Stack and OpenSearch from a capability perspective are multiplying. Thanks in large part to the removal of spyware, code edits, and a host of bug fixes by contributors, OpenSearch Beta is hitting the ground running with real promise for improvement over Elasticsearch. Yet, there is a HUGE difference in the divergent licensing schemas.
While the SSPL and Elastic License v2 prevent adoption of the two projects into other products – in many ways the very lifeblood of their existence. The introduction and evolution of OpenSearch will continue across the community of contributors, of which Logz.io is an enthusiastic participant.
A Head-to-Head Comparison: OpenSearch vs Elasticsearch
OpenSearch
Elasticsearch
Cloud-Agnostic
Yes
Yes
Free to use
Yes
Yes
Aggregations
Yes
Yes
Free to adapt for services
Yes
No
Free to adapt for products
Yes
Depends
License
Open Source (Apache 2.0)
ELv2, SSPL
Version
Fork of Elasticsearch 7.10; OpenSearch Beta (0.0.1)
Elasticsearch 7.11
Query Language
Multiple
Elasticsearch DSL; Kibana QL (KQL)
Internal Plugins
Yes
Yes
External Plugins
Yes
Yes
Will this shift adversely affect project maturity?
Quite the contrary, as we believe that this shift will reignite activity around these essential open source projects which have unfortunately been distracted by commercial interests. As Logz.io CEO and Co-founder Tomer Levy noted in his related blog post: It’s time for a better path forward for everyone, not just the few.
For example, we at Logz.io are subject-matter experts when it comes to the ELK Stack. That expertise, along with the expertise of many other contributors, will now be contributed back directly into the advancement of OpenSearch and consequently into our own Logz.io products.
We’re thrilled that Logz.io will be based on OpenSearch in the near future. We are confident that it will outpace Elasticsearch, offering more advanced features than ever before.
We firmly contend that the legion of collaborators contributing to the OpenSearch project, many of whom have a similar level of expertise, driven by an independent view of where improvements can be made, will result in the best direction possible for this critical software effort.
So what does this really mean for Logz.io?
Here at Logz.io, our own product development work has led to an advanced codebase for operating Elasticsearch and Kibana. We have created, maintain, and continue to update a series of assets such as blogs, user guides, webinars and more covering best practices and tutorials for the ELK Stack. But our expertise goes beyond that, and that work will hardly go by the wayside under OpenSearch.
Is Logz.io using an obsolete search feature?
In a word, no. Logz.io has always adapted Elasticsearch and Kibana into its own platform, making numerous enhancements to the code. That is no different at the present time using OpenSearch – we will continue to maintain the platform that incorporates the last version of non-SSPL-ELv2 Elasticsearch and Kibana (version 7.10) and incorporate our own changes and features into what had been open source code under that version.
What can I do to get involved in OpenSearch?
As Tomer and others both inside and outside of Logz.io have noted publicly, we’re both excited to play a leading role in this endeavor, and just as eager to work with the broader open source community to iterate, enhance and uplift this essential technology.
