As the number one enterprise search engine, any news regarding Elasticsearch is promised to generate a buzz. After the recent release of a second alpha, anticipations and predictions about the upcoming version of ElasticSearch 6.0 are at an all-time high. With new features across the entire stack, this update is significant, and users are clamoring for information and insight into what they can expect from this significant ELK upgrade.
If you are an avid user searching for information on Elastic’s latest offering, you came to the right place. While ELK’s 6.0 improved version provides users with new features in ElasticSearch, Kibana, and Logstash, in this post, we’ll focus solely on ElasticSearch. Read on to learn what to expect from the latest version of this wildly popular engine.
As Elastic Stack’s foundation and flagship product, it makes sense that ElasticSearch is subject to the most noteworthy changes. These updates include doc values, index sorting, and architectural and performance modifications. In the following sections, we’ll break down what the ElasticSearch changes entail.
Issues and challenges with version 5.0 have been well-documented, and ElasticSearch 6.0 addresses many of these. The new Rolling Restart feature eliminates the need for a full cluster restart and allows for increased efficiency. Tribe Nodes have been replaced in 6.0 and can be incorporated in cross-cluster searches. This latest iteration of ElasticSearch also includes improved deprecation logs. The updated logs are more user-friendly and provide users with valuable information guides to illustrate and explain the modifications.
Faster searches are always well-received, and that is one of the key features that the new ElasticSearch offers. Whereas previously sorting was performed during a search, ElasticSearch 6.0 allows for sorting to co-occur when indexing. A key throughput saver is the new version’s ability to terminate search requests sooner.
While index sorting is sure to enhance search efficiency, it is important to note that it is not ideal for all search scenarios. For example, searches with aggregations would not be well-suited for index sorting.
Why are mapping types gone, you ask? Great question. Removing this feature makes it easier for users to comprehend and utilize the foundational data structure in ElasticSearch. Indices in 6.0 will have only one type, not multiple, and a new field has been introduced to store the relationship between indexed documents.
According to ElasticSearch, if a node previously disconnected from the cluster due to a network problem or restart, “each shard on the node would have to be resynced by comparing segment files with the primary shard and copying over any segments that were different.”
A new feature, dubbed “Sequence Numbers,” promises more effective and efficient shard recovery. How does it work? An ID is assigned to each index, update, and delete operation. The IDs are logged in the primary shard’s log of transactions, allowing a copy to reference a log’s processes and use those actions to perform an update, rather than duplicating every file. This feature means faster recovery and improved efficiency.
Sparse Doc Values is a new encoding format designed to prevent the unnecessary utilization of disk resources. Let’s flesh that out by illustrating a sparse values situation. For example, documents lacking values for the fields in the indices tie up a significant amount of disk space and file-system cache. Luckily, Lucene 7 and ElasticSearch 6.0 support Sparse Doc Values, serving as an ally in your war against waste.
According to Elastic, 2236 pull requests by 333 committers have been added since the release of ElasticSearch 5.0.0.
The features and changes outlined above, and implemented based on Lucene 7.0.1, signal another significant release from Elastic. With fixes specifically designed to address issues encountered in 5.0 and new features and functionality, ElasticSearch 6.0 is sure to raise the bar yet again and continue to set the standard for the industry.
Keep an eye on our website for more updates and news on Elastic Stack and the new ElasticSearch 6.0. We’ll continue to publish insights and some key takeaways as new offerings are available.
If you’re interested in more, be sure to send us an email at [email protected]! And get in touch with Weblink if you have tough Elasticsearch problems you need help with!
Weblink Technologies, a leader in Elasticsearch products, provides a solution based solely on Elastic-search. As an Elastic partner and reseller, we have worked with many of customers by providing expert consulting and implementation for Elasticsearch, Logstash, Kibana (ELK), and Beats. Whether you are using Elasticsearch for a web-facing application, your corporate intranet, or a search-powered big data analytics platform, our Elasticsearch experts bring end-to-end services that support your search and analytics infrastructure, enabling you to maximize ROI.
Contact us at [email protected] to learn more about how we can help you leverage Elastic products for high-performing, easy-to-maintain, and scalable search and analytics solutions.
