August 2020 Summaries

YugabyteDB participated in KubeCon + CloudNativeCon Europe 2020 as a sponsor and helped keep #KeepCloudNativeConnected by donating over $3,000 to charity through virtual swag redemptions. YugabyteDB is an open source, high-performance distributed SQL database built on a scalable and fault-tolerant design inspired by Google Spanner, offering PostgreSQL wire compatibility, cloud native features, and advanced RDBMS features like stored procedures and triggers. Popular resources from the virtual booth include blog posts and videos on building cloud-native SQL apps with low latency, monitoring YugabyteDB with Prometheus and Grafana in Kubernetes, and getting started with Hasura GraphQL and YugabyteDB on GKE. The Distributed SQL Virtual Summit is a free online conference to be held September 15-17, featuring thought leaders and innovators discussing the benefits of leveraging distributed SQL systems to build cloud-native applications and services.

To get the total row counts of data in tables across various dimensions (per-table, per-schema, and in a given database), one can use SQL techniques that are similar for PostgreSQL and YugabyteDB. A user-defined function (UDF) named `count_rows_of_table` can be created to count the number of rows in a single table. By querying the `information_schema.tables` table and using this UDF, row counts can be obtained per-table, per-schema, and across all tables in a database. Additionally, aggregate row counts can be calculated by grouping the results by schema or by summing up the individual row counts. These techniques are useful for sanity-checking data after migrating an application from PostgreSQL to YugabyteDB and can help prevent SQL injection attacks when generating dynamic SQL queries.

This summary provides an overview of how Envoy Proxy's PostgreSQL & TCP filters can be used to collect Yugabyte SQL statistics. The process involves installing Docker, creating a configuration file for Envoy, building the containers, and running them in a docker-compose stack. The PostgreSQL filter captures runtime statistics from the database server while remaining transparent to it, providing features such as decoding session information, capturing transaction information, and exposing counters for different types of statements. The TCP filter provides network-level statistics relevant to YSQL, including bytes transmitted and received. The setup allows developers and operations engineers to collect SQL statistics without deploying or managing additional software, making it a convenient integration for YugabyteDB deployments.

Hasura GraphQL and Distributed SQL are technologies used for building scalable and real-time applications. The YugabyteDB Community Q&A session covered GraphQL fundamentals and provided resources for further exploration of these technologies. Upcoming events include a meetup on extending PostgreSQL to Google Spanner architecture, SpringOne connecting Spring apps to distributed SQL clusters in Kubernetes, and the Distributed SQL Virtual Summit 2020 featuring speakers from prominent companies like Pinterest and Twitter. The summit will take place on September 15th-17th, offering insights into scaling GraphQL subscriptions with YugabyteDB.

Getting started with Hasura GraphQL remote JOINs on multi-cloud distributed SQL involves setting up two YugabyteDB clusters, one on Google Kubernetes Engine (GKE) and another on Azure Kubernetes Service (AKS), each with a Hasura GraphQL Engine attached. The setup includes creating databases, tables, and data, as well as configuring a Remote Schema and Remote Relationship to join data from the two databases into a single result set. With this configuration, users can issue GraphQL queries that process data located in two different YugabyteDB databases, hosted on two different clouds, providing a scalable and highly-available GraphQL API.

Deploying a real-time location app with Hasura GraphQL Engine and Distributed SQL requires installing a 3-node YugabyteDB cluster on Google Kubernetes Engine (GKE) and deploying the Hasura GraphQL Engine on GKE. The application uses React and leverages Hasura's real-time capabilities using subscription to track vehicle locations in real-time, powered by a 3-node YugabyteDB cluster. Distributed SQL databases are gaining popularity for their scalability and high availability features, and YugabyteDB is an open-source, high-performance distributed SQL database built on a scalable and fault-tolerant design inspired by Google Spanner. By combining Hasura GraphQL Engine with YugabyteDB, developers can create a scalable and real-time location tracking application that leverages the strengths of both technologies.

Spring Data makes it easy for Java developers to use data access technologies, relational and non-relational databases, map-reduce frameworks, and cloud-based data services. When combined with YugabyteDB, a distributed SQL database that offers benefits such as geo-data distribution, high performance, and horizontal scalability, Spring developers can leverage their familiarity with PostgreSQL while gaining the added benefits of distributed SQL. This tutorial demonstrates how to get started with Spring and YugabyteDB by installing a 3-node YugabyteDB cluster on Google Kubernetes Platform, deploying a REST version of the Spring PetClinic sample application, configuring an Angular frontend for the application, basic testing using Swagger and the Angular UI, and running the application. The tutorial shows that developers can create a dedicated database called petclinic and set a password, install and configure the REST version of Spring PetClinic, run the PetClinic application, configure an Angular frontend, simple data manipulation in Angular, and simple data manipulation in Swagger.

I spent my time as a software engineering intern at Yugabyte, working on various projects and learning from experienced engineers. I was initially introduced to the company through technical due diligence under Partner & CTO Bhaskar Ghosh, where I gained high-level knowledge about the NewSQL ecosystem. After completing interviews with engineers Bogdan and Neha, who led the DocDB and YQL teams respectively, I started my internship in May, working on a project called "Tablegroups" that aimed to enable users to define groups of tables whose data would be stored together on the same node(s). Through this experience, I gained hands-on engineering skills, learned about Kubernetes deployments and cloud-agnostic software support, and contributed to a complex C/C++ codebase. The project involved extending the grammar, implementing the query layer, handling storage implementation, and addressing various roadblocks. I also explored how users might leverage Tablegroups and their future potential, particularly in combination with row-level geo-partitioning. Overall, my experience at Yugabyte solidified my passion for distributed systems & databases and provided valuable insights into software development and product development processes.

Here's a neutral, objective, and interesting summary of the text in one paragraph: YugabyteDB provides various features and tips for developers, including handling arbitrary precision numbers, displaying server IP addresses in the UI, renaming large tables under high load, and creating secondary indexes. The database inherits PostgreSQL column types and allows for transactional metadata changes when renaming tables. Secondary index rows are stored internally similarly to main table rows, with non-unique indexes having a primary key consisting of multiple columns and unique indexes having a single primary key column. To display server IP addresses in the UI, set the `--webserver_interface` flag. YugabyteDB has new documentation, blogs, tutorials, videos, and upcoming events, including KubeCon + CloudNativeCon Europe 2020 Virtual and GraphQL-related meetups. The database provides a quickstart page for getting started with local development.

KubeCon + CloudNativeCon Europe 2020 is a virtual event taking place August 17-20, where various open source and cloud native communities will come together to drive cloud native computing forward. YugabyteDB, an open source high-performance distributed SQL database, will be showcased at the event, featuring its PostgreSQL wire compatibility, cloud native design, and support for advanced RDBMS features. The company invites attendees to visit their virtual booth to connect with them, learn about distributed SQL, cloud native application development, Kubernetes, and more. YugabyteDB is committed to open source, running on various Kubernetes platforms and integrating with multiple open source projects. Attendees can also redeem a virtual swag for charity, while learning more about the company's commitment to open source through various resources.

yugabyted is a native server that acts as a parent server across the YB-TServer and YB-Master servers, removing the day-1 learning curve for new users to experience YugabyteDB's benefits in the shortest time possible. It allows users to create single-node clusters easily, review cluster status, connect with ysqlsh and load sample data sets, run queries, and simulate multi-node deployments across multiple host machines. The `--join` option enables adding nodes to an existing cluster, enabling horizontal write scalability and self-healing architecture that elects new shard leaders on available nodes whenever shard leaders are lost due to node failures or network partitions. yugabyted is currently in Beta and welcomes user feedback via GitHub and Slack as it continues to advance its feature set over the next few releases.

GraphQL & Distributed SQL Tips and Tricks – Aug 10, 2020` This blog post provides tips and tricks for combining GraphQL with YugabyteDB, a distributed SQL database, to develop scalable APIs and services. It covers various use cases, including microservices and mobile apps, and explains how to integrate Hasura, an open-source GraphQL project, with YugabyteDB. The post also discusses how to create computed fields, event triggers, and row-level permissions using Hasura and YugabyteDB. Additionally, it highlights new documentation, blogs, tutorials, and videos, as well as upcoming events and a quickstart guide for getting started with YugabyteDB.

Flyway is an open source database version control and migration tool that supports multiple databases including Oracle, SQL Server, DB2, MySQL, PostgreSQL, and YugabyteDB, which is compatible with PostgreSQL wire protocol. Flyway provides seven commands for managing database version control: Migrate, Clean, Info, Validate, Undo, Baseline, and Repair. The tool can be used to deploy and roll back schema changes to a distributed SQL database like YugabyteDB using the PostgreSQL JDBC driver. A 3 node YugabyteDB cluster was installed on Google Kubernetes Platform (GKE) for this demo, with Flyway installed locally and connected to the YSQL service of the cluster. The tool was used to run database migrations against the YugabyteDB database, including creating a schema history table and inserting data into the motorcycle_manufacturers table. Additionally, the `clean` command was used to drop all objects in the configured schemas.

Distributed SQL in YugabyteDB can be queried using the WriteTime function to retrieve timestamps, per-row on-disk space usage can be checked by loading data into a table and then flushing it to disk, JSONB fields can be searched for property names using the `->>` operator, query logging can be enabled using the ysql_log_statement flag, and new documentation has been published to support the release of YugabyteDB 2.2. Various upcoming events and resources are also available for users to learn more about YugabyteDB.