March 2020 Summaries

This new version of Index Advisor service provides several key improvements and fixes, including support for TTL (time-to-live) metadata, index recommendations on covering indexes, adjusting the order of index keys for functional and array predicates, and more. The service is now available to users every month, with fixes and recommendations made available before the next maintenance release. The Index Advisor service aims to help database administrators, developers, and architects optimize query performance and meet SLAs by providing index recommendations without requiring users to create a bucket or upload schema or data. Users can find this service useful if they want to avoid reading index creation rules, need help creating indexes for their queries, or want to generate advice on indexes without downloading the latest Couchbase server version.

MongoDB and Couchbase are two popular NoSQL database systems that have evolved their replication solutions to address the growing need for highly available, scalable, and globally distributed deployments. MongoDB's replication architecture is based on a replica set, where each secondary copies data from the primary asynchronously, but can become a single point of failure if not managed properly. In contrast, Couchbase follows a peer-to-peer architecture, allowing multiple independent clusters to be created and replicated across different geographies. This provides flexible topology, workload isolation, and heterogeneous scaling benefits. Both systems offer cloud deployments with MongoDB's Global Cluster and Couchbase's cross-datacenter replication solution, which support true active-active setups for concurrent writes across multiple regions. While MongoDB's setup can become complicated as it scales, Couchbase's deployment is considered more intuitive and easier to manage. Additionally, Couchbase offers advanced filtering to replicate subsets of data and prioritizes ongoing replication compared to new replication or vice versa.

Couchbase Full Text Search (FTS) offers a more efficient alternative to Global Secondary Index (GSI) for indexing and querying multiple arrays, as demonstrated using the Travel Sample dataset. The example involves searching hotel documents based on public likes and reviews using N1QL queries, encountering limitations with GSI's inability to index multiple arrays in a single query due to b-tree index restrictions. By creating separate array indexes with Couchbase N1QL Index Advisor, the queries run successfully but inefficiently, fetching excessive documents. In contrast, FTS allows for a single index on multiple arrays, significantly improving performance by returning the necessary documents faster with reduced resource usage. Utilizing the SEARCH() function in N1QL, the FTS solution returns the same results in a fraction of the time, showcasing the integration's capability to enhance query performance and system efficiency by overcoming GSI's limitations.

Transitioning from Oracle's relational model to Couchbase's JSON model involves careful consideration of data type mapping, as Oracle requires explicit declaration of column types, while Couchbase automatically interprets data types through JSON syntax. The conversion encompasses various data types, including strings, dates, booleans, numerics, binary data, and large objects, each with specific considerations for mapping. For instance, in Couchbase, string data is inherently Unicode with a maximum length of 20MB, and date and time are represented in ISO 8601 format as strings, offering extensive functions for formatting and arithmetic. Oracle's complex data types such as objects and arrays have built-in support in Couchbase's JSON model, allowing for nested levels and versatility in data representation. Additionally, the conversion from primary keys in Oracle to document keys in Couchbase requires careful handling to ensure uniqueness within a bucket, often by converting parts to strings and using separators.

The article delves into the intricacies of handling case sensitivity in databases, particularly focusing on SQL, JSON, and N1QL. While SQL is generally case insensitive to identifiers and keywords, it remains case sensitive to data, whereas JSON and N1QL are case sensitive to both field names and data. It highlights the challenge of performing case-insensitive queries and suggests using functions to transform data into a uniform case, which facilitates efficient querying through index creation. The discussion extends to complex expressions and array functions, emphasizing the importance of optimizing queries for performance by creating appropriate indexes with the help of tools like Index Advisor. Additionally, it explores the use of Full-Text Search (FTS) for more effective text processing and querying, noting the benefits of custom analyzers to ensure comprehensive case-insensitive searches. The article underscores the significance of verifying query plans to confirm that predicates are properly pushed to index scans, thereby enhancing query performance.

Couchbase has implemented a Business Continuity Plan to address the COVID-19 pandemic's impact on its customers and operations. The company is taking steps to ensure continuity, including adjusting business operations, maintaining high-quality services, and supporting employees working from home. Couchbase has canceled or postponed in-person events, minimized business travel, and closed offices to minimize employee exposure while practicing social distancing. Despite the challenges, the company remains committed to its customers and partners, with a promise to maintain support and service levels.

The text provides a detailed guide on configuring network connections for Couchbase clusters using Cross Data Center Replication (XDCR) in a cloud environment, specifically focusing on Amazon's Elastic Kubernetes Service (EKS). It involves deploying Kubernetes clusters in two separate regions, Virginia and Ohio, and setting up Couchbase clusters in each. The guide walks through the process of establishing VPC peering for network connectivity between the clusters, updating routing tables, and configuring security groups to allow communication. It also covers the setup of XDCR to enable data replication from a source bucket in the Virginia cluster (Blue) to a target bucket in the Ohio cluster (Green). The steps demonstrate how to manage network configurations and Couchbase deployments on AWS, though the principles can be applied to other cloud platforms.

The latest preview of the Couchbase Autonomous Operator 2.0 beta introduces several new enterprise-grade features with fully autonomous capabilities, including security, monitoring, high-availability, and manageability. This release builds on features in the Kubernetes platform at its core, re-architecting how it installs, upgrades, and manages Couchbase resources. It includes a new custom resource model, improved user and RBAC management, automated backup management, Prometheus metrics collection, certificate authentication using mutual TLS support, XDCR management, and tutorials on deploying Couchbase Mobile Integration.

The article features interviews with two female engineers at Couchbase, Cecile Le Pape and Mary Roth, who share their inspiring stories of becoming engineers despite the male-dominated field. Both women highlight the importance of overcoming obstacles such as self-doubt and societal expectations to pursue a career in engineering. They emphasize the value of embracing challenges and learning from others to succeed in the field. Cecile notes that she was drawn to engineering due to its creative potential, while Mary credits her mathematical aptitude with driving her interest in computer science. The women offer advice to aspiring female engineers, including being prepared to debate and earn respect on merit, and making a conscious effort to pursue their passions. Their stories serve as an inspiration to others who may be considering a career in engineering.

Organizations transitioning from traditional relational databases to NoSQL document databases face challenges such as converting RDBMS schemas and learning new APIs. The term "NoSQL" is often misunderstood, as it does not imply complete abandonment of SQL; many NoSQL systems still utilize SQL-like query languages. Couchbase's N1QL, for example, extends SQL to support hierarchical data models while maintaining familiar SQL constructs. Altoros, a company assisting in IT transitions, has published a report comparing query languages like MySQL/SQL, Couchbase N1QL, and MongoDB, focusing on criteria such as simplicity, readability, and expressiveness. The report highlights that N1QL closely resembles SQL, making it easier to adapt for those familiar with traditional databases, whereas MongoDB often requires more lines of code and server trips, especially for complex queries. The report, which assesses these languages across nine typical CRM query scenarios, suggests that while Couchbase offers SQL-like ease, MongoDB's approach demands more complex client-side handling due to its limited support for complex aggregations.

The Couchbase Query and Analytics services are two options for querying documents using SQL. The Query service is used for data manipulation within application logic, while the Analytics service is used for reports, analysis, and dashboards. The Query service is more efficient for short, operational queries that retrieve or manipulate smaller amounts of data, whereas the Analytics service is better suited for longer, complex queries that process large amounts of data. Key differences between the two services include their ability to support SELECT, INSERT, UPDATE, DELETE, MERGE operations (Query) and only SELECT operations (Analytics). The Query service performs well on operational queries with simple filters, while the Analytics service excels at more complex queries involving aggregations, joins, and window functions. In a 3-node Couchbase cluster environment, the Analytics service outperforms the Query service in most cases, especially for queries that require joins or aggregations. However, the performance gap narrows when adding Window functions to the query. Understanding when to use each service is crucial for optimizing data access and analysis in Couchbase applications.

Couchbase has launched a developer documentation survey, aiming to gather feedback from all levels of Couchbase users to enhance their documentation. Open to both longstanding customers and new developers, the survey offers participants a chance to win prizes, with four randomly selected winners, subject to legal and geographical restrictions. Participants who complete the survey by March 31st, 2020, also have the opportunity to engage in a live, interactive documentation exercise for a guaranteed reward of $200 upon successful completion.