April 2019 Summaries

Elastic has launched the #ElasticStories hashtag to encourage its community to share their innovative and imaginative uses of the Elastic Stack, aiming to highlight stories ranging from simple implementations to complex use cases. This initiative provides a platform for users to share experiences related to machine learning, security, application search, and other features of the Elastic Stack on Twitter, fostering an exchange of ideas and solutions. By using the hashtag, community members can include links, artwork, or screenshots to illustrate their stories, whether they involve overcoming challenges or reaching milestones. Elastic plans to amplify these contributions by retweeting them from the official @Elastic account and possibly featuring them in their blog, community newsletter, or website, underscoring the belief that community feedback is essential for the continuous evolution of their products and services.

Kuniyasu Sen's blog post discusses the importance of using appropriate language analyzers and dictionaries for effective full-text search in Elasticsearch, particularly for CJK (Chinese, Japanese, and Korean) languages. The default standard analyzer is not suitable for these languages, necessitating the use of specific language analyzers like the Japanese Kuromoji, Korean Nori, and Chinese IK analyzers. These analyzers rely on dictionaries to determine word tokenization, allowing for meaningful search results, as seen in examples like the Japanese word for Skytree. Dictionary updates play a crucial role in ensuring accurate tokenization, affecting both indexing and search queries. To apply dictionary updates to existing indices, Elasticsearch requires reindexing, which can be facilitated using the Update By Query API. While updates may or may not impact search results, understanding these processes enables better search experiences, and the blog offers guidance on implementing and updating dictionaries within Elasticsearch.

Support for Microsoft SQL Server has been introduced to Metricbeat, a tool used to capture database metrics, as part of its generally available modules. This addition includes two new beta metricsets—performance and transaction_log—alongside curated dashboards in Kibana to help users navigate SQL Server metrics. The performance metricset tracks crucial indicators like user connections, transactions, and cache hits, important for assessing server efficiency. Meanwhile, the transaction_log metricset provides insight into server log space usage and backup times, aiding in diagnostics and performance understanding. Setting up the SQL Server monitoring module is made user-friendly through Kibana's "Add Data UI," allowing users to configure and activate the module and access dashboards for detailed database views. This enhancement aims to streamline the monitoring process for Microsoft SQL Server users, encouraging them to deploy the Metricbeat module for improved server oversight.

Monitoring MySQL, Percona Server, and MariaDB databases is essential for maintaining application performance and identifying potential issues, and the Elastic Stack offers tools such as Metricbeat and Filebeat to facilitate this. Filebeat for MySQL includes filesets for analyzing database error and slow query logs, providing insights into server problems and resource-heavy queries, while Metricbeat collects key metrics from the SHOW GLOBAL STATUS query to monitor server health. These tools support MySQL, Percona Server, and MariaDB, offering compatibility with various database versions, including MySQL 8.0 and MariaDB 10.2 and 10.3, and have expanded to include experimental support for Galera clustering metrics. The Elastic Stack provides additional capabilities like Elastic APM for application performance monitoring and Packetbeat for network connection analysis, enabling comprehensive monitoring solutions for database administrators.

Apache Lucene 8 introduces a range of enhancements aimed at improving query execution speed, indexing efficiency, and memory management. Notable features include a new API that allows users to bypass unnecessary counting for queries, thereby speeding up execution by providing a lower bound on the number of matching documents. The update also introduces indexing impacts, which enable maximum score calculations for document blocks, and a new FeatureField for efficient custom-scoring queries. Additionally, Lucene 8 enhances docvalue access speed with jump tables and allows terms dictionaries to be loaded off-heap, reducing heap memory usage. These improvements aim to optimize performance, especially for large document sets and frequent updates, with further enhancements planned for future releases.

In an exploration of enhancing the monitoring of NATS messaging systems, a team in the telecommunications sector, working with Kubernetes microservices, developed NATS modules for Metricbeat and Filebeat to integrate seamlessly with their existing EFK stack. This initiative began during an internal hackathon, where the team quickly created a NATS Beat, which was later incorporated into the core Beats project with support from the Elastic Beats team. NATS, known for its ability to handle millions of messages per second, provides key monitoring data through HTTP endpoints and server logs, which the team utilized to create a comprehensive monitoring solution. This solution includes pre-built dashboards for visualizing metrics and logs, allowing users to gain insights into the performance and state of their NATS environments. The development process not only solved the team's monitoring challenges but also fostered collaboration with Elastic engineers and contributed to the open-source community.

CTcue has developed a privacy-focused platform using the Elastic Stack to enhance the searchability of electronic health records (EHRs), which are predominantly composed of unstructured data that is challenging to process. Leveraging Elasticsearch since 2015, CTcue has optimized its data model to improve search performance by using a denormalized structure that streamlines query processes and mitigates the limitations of on-premises infrastructure. The platform aids healthcare professionals in conducting prospective feasibility studies and real-world evidence integration, facilitating more efficient healthcare delivery and decision-making. As CTcue expands its reach across more hospitals, it is focusing on infrastructure scalability, centralized monitoring, and reporting, while planning to utilize Kibana and Canvas for enhanced data visualization and real-time monitoring. The platform's integration of natural language processing helps refine search results, making unstructured medical data more accessible, with the ultimate aim of improving patient care and internal hospital processes.

User annotations in Elasticsearch, introduced from version 6.6, provide a method for enhancing machine learning jobs with user-specific domain knowledge, helping to better interpret anomalies detected in data. These annotations can be applied to various datasets, such as weather sensor data, to highlight significant events or validate machine learning outputs against historical data. The annotations can be managed through the Single Metric Viewer, which allows users to create, edit, or delete annotations and share them with others via permalinks. Additionally, annotations are stored in a dedicated Elasticsearch index, making them accessible for automated processes, including programmatically creating annotations using Elasticsearch APIs. The integration with Watcher allows users to create curated alerts based on these annotations, sending notifications to platforms like Slack. This feature not only assists in anomaly detection but also improves alerting precision, enhancing the overall utility of machine learning applications in Elasticsearch.

The guide explores how to instrument a Go application using the Elastic APM Go agent to gain insights into application performance and visibility for distributed workloads. It highlights the challenges of instrumenting Go due to its unique compilation and threading model and provides detailed steps on how to instrument a Go app to capture detailed response time performance data, infrastructure, and application metrics. The guide also covers tracing web requests, SQL queries, custom spans, outgoing HTTP requests, and panic tracking, illustrating how to integrate these elements with Elastic APM for enhanced observability. It further explains how to configure the Elastic Stack to receive and visualize data, with instructions on setting up and using Filebeat for logging integration. The article emphasizes the integration of application logs with APM trace data using popular logging frameworks and discusses the importance of capturing errors and metrics to build a comprehensive observability solution.

Auditbeat is a popular Beat that gathers data from the Linux audit framework to monitor processes on Linux systems, providing insights into security-related information, file integrity, and process data. Recently, machine learning job configurations have been introduced for the Auditbeat auditd module, enabling automatic detection of suspicious activities in server kernels or Docker containers. These analyses help identify anomalous user access or errant processes. Users can configure machine learning jobs that analyze rare process activity and high process rates, which are crucial for spotting potentially malicious activities hidden among common processes. The module offers dashboards, visualizations, and saved searches for both on-host and Docker environments, allowing for detailed investigations into identified anomalies, such as rare processes or unusual spikes in process activities, which might indicate security threats like flooding-style attacks.

Kibana's April 2019 update highlights the release of version 7.0 and ongoing efforts toward future developments, including hiring for various engineering and management roles within the team. Key advancements include the integration of security feature control, improvements to the Elastic-Charts functionality, and enhancements in data visualization and reporting, such as the ability to sort series and reduce resource consumption in queue processing. The platform is preparing for plugin infrastructure enhancements, and the Canvas layout engine has undergone significant refactoring to allow for dynamic data-driven layouting and improved filter functionality. The team is also advancing TypeScript conversion based on developer feedback and is working on design documentation and styling for Elastic Charts in EUI, aiming to enhance user experience and provide design guidelines. Additionally, the update features notable progress in the secret service and introduces new features in the Geo-Maps app, with a focus on enhancing geospatial analysis capabilities.

Since its introduction in June 2017, Elastic APM has evolved into a comprehensive application performance monitoring solution, continually enhancing its features to better serve developers and operations teams. The APM UI, designed with user-friendly dashboards, allows for efficient data analysis and debugging. The design process, which began amidst a major redesign of Kibana and the introduction of the Elastic UI (EUI), involved establishing core components and integrating custom designs like the Transaction Timeline. Over time, the design team has focused on researching user needs and iteratively developing features, such as distributed tracing and metrics, while adapting the UI to accommodate new functionalities and user scenarios. The team's approach includes creating wireframes, high-fidelity mock-ups, and eventually implementing features, often releasing them incrementally to gather user feedback and make improvements. With each release, from the initial general availability to the integration of machine learning and distributed tracing, Elastic APM has incorporated feedback to refine its offerings and support a seamless user experience. The ongoing development aims to enhance the existing features and introduce new capabilities, inviting users to contribute feedback and encouraging designers to join the team for future innovations.

Liselot Poppink, Consulting Director for Europe, Middle East, and Africa at Elastic, shares her journey and experiences as a woman in the tech industry, emphasizing the importance of freedom and responsibility in the workplace. Initially, Poppink faced challenges and skepticism in her studies and early career due to gender biases, but she persevered and transitioned from programming to consulting, drawn by the opportunity to engage directly with clients and improve IT communication. She joined Elastic after being attracted by its culture of transparency and growth, where she now leads a consulting team that helps clients maximize the value of Elastic's products. Poppink highlights the supportive environment at Elastic, which values innovation and personal growth, and she advises women in tech to remain confident, seek feedback, and pursue what energizes them.

The release of Elastic Stack 7.0 marks a significant advancement in observability, providing enhanced visibility into complex application infrastructures through new and improved modules. Key features of this release include four new modules, such as the NATS module for Kubernetes and Cloud Foundry, a beta Microsoft SQL Server module, a CouchDB module, and a highly anticipated AWS EC2 module for Cloudwatch metrics. Additionally, the Prometheus module has undergone substantial improvements to better integrate time-series metrics with Elasticsearch, allowing for seamless comparison and correlation with logs, application traces, and other data. The adoption of the Elastic Common Schema (ECS) further facilitates data correlation and analysis across diverse sources, ensuring a consistent approach to data modeling. These updates collectively empower users to monitor their application stacks comprehensively, from infrastructure components to business data, enhancing their ability to quickly identify and address issues.

Elasticsearch 7.0.0, based on Lucene 8.0.0, introduces significant enhancements in speed, scalability, and usability, positioning it as the most efficient and resilient version to date. Key improvements include performance enhancements for search and indexing, such as faster top k retrieval and adaptive replica selection, which optimizes query speed and search throughput. The new version also simplifies cluster coordination, improves memory management with a real-memory circuit breaker, and reduces oversharding by defaulting to one shard per index. Usability is enhanced with bundled Java, JSON logging, and a feature-complete high-level REST client for Java. Additionally, Elasticsearch 7.0.0 supports cross-cluster replication by default and introduces new field types and query capabilities, such as nanosecond timestamps and intervals queries, to broaden its application scope and improve user experience.

The release of Elastic Cloud Enterprise (ECE) 2.2.0 brings significant enhancements, focusing on improved security, management, and scalability in multitenant environments. Key features include a cross-cluster search UI for simplified management of multiple clusters, role-based access control in beta for fine-grained user management, and integration with index lifecycle management for automated index operations. Additionally, the update introduces Elasticsearch keystore support for secure settings storage, new Ansible playbooks for easier installation and management, and readiness for Elastic Stack version 7.0 with support for zero-downtime major version upgrades. Performance improvements, such as more efficient use of ZooKeeper, further enhance the platform's scalability and usability.

The article explores the implementation of a hot-warm-cold architecture using Elasticsearch's Index Lifecycle Management (ILM), a feature designed to efficiently manage data indexes. This architecture is particularly suited for time series data like logs and metrics, where data is categorized into hot, warm, and cold phases based on its usage and importance. The hot phase contains the most accessed data, requiring high CPU and fast I/O, while the warm and cold phases store less frequently accessed data, needing more disk space but less processing power. ILM allows users to define and automate the movement of data between these phases, optimizing costs and performance. It includes actions like rollover, force merge, and freeze to manage data efficiently, with policies tailored to specific needs. The article also highlights configuring ILM policies for Beats and Logstash, demonstrating how ILM simplifies data management without needing external tools like Curator. With Elasticsearch version 7.0, ILM is enabled by default for Beats and Logstash, further streamlining the process.

Elasticsearch has moved towards the removal of types in its system, which began with version 5.0 and has been progressively implemented through versions 6.0, 7.0, and 8.0. Originally, types were intended to provide multi-tenancy within a single index, but they proved to be problematic and incompatible with Lucene. To ease the transition for users, Elasticsearch introduced changes such as enforcing compatible mappings for fields with the same name across different types, preventing multiple types in new indices, and deprecating APIs that accept types. The transition involves adopting new "typeless" APIs, which no longer require or support types in URL paths, request bodies, or response bodies. Users are advised to perform certain upgrades, such as moving to version 6.8 before 7.0, stopping the use of _default_ mappings, and making adjustments to API calls to accommodate these changes. The ultimate goal is to eliminate the use of types entirely by version 8.0, requiring users to adapt their systems to the new typeless structure.

Logstash 7.0.0 introduces significant enhancements for users, featuring a default Java pipeline execution, automatic index lifecycle management (ILM), and updates to plugins. Originally built in Ruby, Logstash now leverages JRuby for improved performance, but recent developments have shifted its internals to Java, allowing Java-based plugins to interact directly with a Java pipeline for better efficiency and reduced memory use. The new Java execution engine compiles directly to JVM bytecode, offering up to 20 times faster startup and reload times for large configurations. The release also includes automatic ILM deployment with the Elasticsearch output plugin, which detects ILM support in the cluster, and allows users to enable or disable this feature. Additionally, deprecated and obsolete settings in various plugins have been addressed, marking a significant update in Logstash's usability and performance.

Kibana 7.0.0 introduces a range of significant updates, including a refreshed design with a global header and a collapsed side navigation for more space, as well as the new default Kibana Query Language (KQL) that simplifies query syntax with features like autocomplete for users with a Basic license or higher. The release also includes a dark theme, responsive dashboards optimized for mobile use, and revamped timepicker and filter functionalities for improved user interaction. The saved objects structure has been enhanced for better export and transfer of dashboards and their dependencies, while Canvas now features an expandable expression editor and additional keyboard shortcuts for improved usability. Overall, these updates aim to enhance user experience and functionality, inviting users to explore the new features and provide feedback through their Discuss forum.

Beats 7.0.0 has been released with significant updates including the adoption of the Elastic Common Schema (ECS), which standardizes document fields for event data in Elasticsearch, facilitating easier data correlation and content development. The release also introduces index lifecycle management (ILM) for more dynamic index handling based on factors like shard size and performance, and recommends using Metricbeat for monitoring the Elastic Stack. Several new modules are introduced, such as the AWS EC2 module for cloud resource metrics, and enhanced support for security analytics data sources, including integrations with Zeek and Santa for macOS. Users are encouraged to upgrade, review the documentation for migration guidance, and give feedback on the new features.

Elastic has announced the achievement of SOC 2 Type 2 and CSA STAR Level 2 Attestation for its Elasticsearch Service, Elastic Site Search Service, and Elastic App Search Service, marking a significant milestone in their cloud security compliance efforts. The SOC 2 Type 2 certification ensures that the services have met the AICPA's Trust Services Criteria, focusing on security, confidentiality, and availability, which evaluates the suitability and operational effectiveness of their controls. Additionally, the CSA STAR Level 2 Attestation, part of a collaboration between the Cloud Security Alliance and the AICPA, provides independent third-party assessments of cloud providers, emphasizing transparency, rigorous auditing, and standard harmonization. These certifications apply automatically to current workloads, enhancing trust and reliability for Elastic's cloud services.

Elastic announced the release of the fourth iteration of the Elastic App Search beta, focusing on enhancements in performance and bug fixes. Notable updates in this beta include improved intelligence and messaging for configuration issues, the introduction of Query Suggestions, refined email aesthetics, and better handling of integers and booleans in the app_search.yml file. Additionally, several behind-the-scenes improvements have been made to optimize the system. Users who have been utilizing previous beta versions are advised to start fresh with this new release, as it is approaching general availability. Feedback from users is encouraged to refine the product further, and a free test drive is available until the general availability launch.

Elasticsearch for Apache Hadoop 7.0.0 has been released, aligning with Elasticsearch 7.0.0 and marking significant updates, including a shift to require Java 8 or higher, as the support for older Java versions is phased out. This release also sees the removal of the Cascading integration, which was deprecated in version 6.7.0 due to declining usage and the decision to focus resources on more beneficial areas for the user base. Additionally, the update simplifies the partitioning process by reverting to using one input partition per shard by default, rather than slicing shards into smaller pieces, which had previously led to confusion and performance issues. This change is aimed at enhancing the user experience in terms of getting started and tuning the system. The release encourages user feedback and suggestions for further improvements through forums and GitHub.

Elastic Stack 7.0 introduces major enhancements across its components, marking a significant release with over 10,000 pull requests from 861 contributors. Kibana 7.0 features a new design with a minimal UI, a global navigation system, dark mode, and responsive dashboards for improved mobile usability. In Elasticsearch, the release includes a new cluster coordination layer for better scalability and reliability, alongside a real memory circuit breaker to enhance node stability. Search capabilities are boosted with faster top k queries using the Block-Max WAND algorithm, simpler intervals queries for proximity-based searches, and an upgraded function score 2.0 for advanced search relevancy control. Elastic Maps benefits from a new geotile grid aggregation for stable zooming, and time series use cases see improvements with nanosecond precision timing, thanks to a shift from JODA to the Java time API in JDK 8.

The Kibana development team made significant progress on various fronts as of April 2019, focusing on platform enhancements, security improvements, and UI refinements. They advanced in integrating feature controls into applications and successfully merged new platform UI plugin services and API documentation improvements. The team also worked on decoupling Angular components and enhancing alerting services, while telemetry data collection intervals were under review to reduce frequency. The Geo-Maps App saw advancements in embeddability and filtering, and the introduction of a new architecture for the Tile Service was completed. The Kibana App received updates, including a new visual editor and improved charting capabilities, while efforts continued to migrate older components to the new platform. Lastly, enhancements were made to the Canvas and Design components, including new drag-and-drop functionalities and UI cleanups for better usability and alignment with EUI standards.

The article, authored by Jess Smith, details the iterative redesign process undertaken by Elastic's design team to create a cohesive visual identity across its diverse and rapidly expanding range of products and solutions. Instead of a superficial rebrand, the effort was a thoughtful collaboration aimed at establishing a scalable design system that harmonizes tech and marketing needs. Initially sparked by the merger with Swiftype, the redesign journey involved addressing inconsistencies in existing logos and icons, which, although conceptually sound, had become insufficient due to the company's growth. The team focused on establishing a clear visual hierarchy between icons and logos, crucial for product recognition and user interface interaction. Through trial, error, and community engagement on open-source platforms, they refined their branding approach, culminating in a cohesive visual strategy that reflects Elastic's flexible and evolving nature. The process underscored the importance of visual strength and brand cohesion, ultimately leading to a successful system that supports ongoing innovation and adaptation.

Elastic and Google have announced an enhancement of their partnership to provide a more native integration of the Elasticsearch Service on Google Cloud Platform (GCP). Having offered the Elasticsearch Service on GCP since 2017, Elastic allows users to deploy the latest versions of Elasticsearch, Kibana, and a variety of features for tasks such as security, machine learning, and geospatial analysis. The expanded collaboration aims to make these features more accessible to GCP customers through the Google Cloud Console, allowing them to manage Elasticsearch alongside other GCP services and benefit from streamlined workflows for management and billing. Existing customers will gain more flexibility in procurement and payment options, while plans are underway to introduce more capabilities and expand the service to the Asia Pacific regions within the year. Additionally, new users can explore the Elasticsearch Service on GCP through a 14-day free trial.

Instrumenting a Ruby application with the Elastic APM Ruby agent is streamlined by installing the elastic-apm Rubygem, which offers built-in support for Rails and Rack, alongside options for custom instrumentation. Users can enhance data collection by assigning specific tags and user information, making it easier to identify performance issues specific to certain customers. The SpanHelpers module allows tracking the duration of specific methods, and a public API enables manual creation of transactions and spans for more granular control. The agent simplifies the initial setup while offering advanced tools for detailed application analysis, allowing users to either run the Elastic stack locally or use the Elasticsearch Service on Elastic Cloud to monitor application performance and optimize user experience.

Elastic has celebrated the opening of its new Amsterdam office in the 5 Keizers Building, marking a significant expansion from its original Schinkelbuurt location established in 2012. The event, which took place in early 2019, included a ribbon-cutting ceremony by Constantijn van Oranje, Prince of the Netherlands, and featured a live AMA with Elastic's founders, Shay Banon, Steven Schuurman, Uri Boness, and Simon Willnauer. This expansion reflects Elastic's rapid growth, highlighted by its IPO on the NYSE in October 2018. The opening festivities also included collaboration with StartupDelta to host ten Dutch startups, promoting a spirit of community and innovation. Attendees had the opportunity to engage with Elastic engineers through product demos and enjoy a celebratory gathering with music, food, and drinks, emphasizing Elastic's commitment to maintaining its unique culture and small office feel despite global expansion.

Cross-Cluster Replication (CCR) in Elasticsearch 6.7 is a feature that enables data replication across multiple datacenters or Elasticsearch clusters without the need for additional technologies. By configuring replication at the index level, CCR supports various strategies, such as data locality and centralized reporting, making it suitable for diverse use cases like maintaining data close to application servers or replicating banking data globally for reporting purposes. The CCR feature is managed through Elasticsearch's APIs and Kibana UI, offering options like auto-follow patterns for time-based indices and defining remote clusters for replication. Security requirements necessitate specific privileges for users in both source and target clusters. CCR is a platinum-level feature that can be trialed for 30 days, and it provides flexibility in architecture, including production and disaster recovery setups, multi-datacenter deployments, and bi-directional replication. The tutorial outlines setting up CCR, creating indices for replication, and testing the setup, while administrative APIs facilitate management tasks like pausing, resuming, or converting a follower index to a normal index.

Hill Enterprise Data Center (HEDC) at Hill Air Force Base utilizes Elastic Cloud Enterprise (ECE) to manage and secure its geo-dispersed data center by aggregating and analyzing data from over 100 information systems. This system supports the U.S. Air Force logistics center's operations, ensuring compliance with National Institute of Standards and Technology (NIST) requirements and Department of Defense (DoD) standards. HEDC employs role-based access control (RBAC) for secure data sharing and access across Portable Operating Databases (PODs) while facilitating efficient data transfer and analysis through Elasticsearch. The implementation of ECE allows HEDC to overcome challenges associated with paper compliance and data silos, providing a streamlined process for data ingestion, correlation, and secure access, which is essential for maintaining operational efficiency and compliance within the DoD framework.

The blog post discusses the integration of Elasticsearch with Prometheus metrics, promoting the adoption of open standards such as OpenMetrics to enhance observability. It emphasizes the importance of open standards for interoperability, vendor neutrality, and maximizing user choice. Prometheus, recognized as a de-facto standard in cloud-native metric monitoring, offers an exposition format that facilitates the collection and visualization of metrics. The blog details how Elasticsearch can ingest Prometheus metrics, combining them with logs and APM data for comprehensive monitoring and analysis using Kibana. It also provides examples of deploying Metricbeat within a Kubernetes environment to scrape Prometheus metrics, demonstrating how to visualize these metrics within the Elastic Stack. The post underscores the value of integrating logs, metrics, and traces to build an observable system capable of identifying and diagnosing production issues effectively.

The migration guide from AWS Elasticsearch Service to Elasticsearch Service on Elastic Cloud outlines a detailed process for users looking to leverage advanced features and support provided by Elastic, which are not available from Amazon. It offers a step-by-step method to perform a "lift-and-shift" migration, starting with a snapshot of the AWS ES cluster stored in an S3 bucket, and then restoring it to an Elastic Cloud deployment. The guide emphasizes the necessity of programming skills and familiarity with AWS SDKs, especially in Python, to successfully complete the migration. Users must set up IAM roles and policies, create necessary AWS resources, and execute Python scripts for snapshotting and restoring. Additionally, the guide notes that Elastic Cloud offers a 14-day free trial, supports multiple cloud providers such as AWS, GCP, and Azure, and provides ongoing updates and unique features like Canvas, APM, and machine learning capabilities.

On Equal Pay Day 2019, Leah Sutton from Elastic highlighted the company's efforts to address gender pay gaps, emphasizing the symbolic significance of the day as a reminder of ongoing disparities. Elastic had previously undertaken a manual review of salaries to identify and correct unexplained differences between male and female employees, ceasing the practice of basing offers on past pay to prevent perpetuating gaps. By 2019, with nearly 1,400 employees globally, Elastic partnered with Economists Incorporated to conduct a comprehensive statistical pay equity analysis. The analysis, involving multiple regression techniques, found no statistically significant pay disparities between men and women, with female employees in the U.S. earning 99.6 cents for every dollar earned by men, and globally earning 98.8 cents. In engineering roles, women earned $1.03 for every dollar earned by their male counterparts. Elastic plans to continue annual reviews to maintain pay equity and hopes to incorporate more demographic data to identify potential gaps among other groups, particularly people of color.

The announcement details the release of the first release candidate for a new Elasticsearch JavaScript client, following a comprehensive refactoring process aimed at resolving versioning issues and improving user experience. The new client will be released under the @elastic npm organization and promises numerous enhancements, such as modern syntax, improved performance, and TypeScript support, offering a more intuitive and consistent developer experience. The package reorganization involves publishing a new scoped package and transforming the existing package into a pointer to the latest version, allowing for a smoother transition for users. The refactoring results in breaking changes, such as the removal of browser support and the introduction of new error handling mechanisms, necessitating code adjustments for users of the previous client version. The release aims to align with the final release of Elasticsearch 7.0, and users are encouraged to report bugs and propose features to further refine the client.