An Overview of the New MySQL 8.0

mysql desktop

MySQL remains a valid database option for companies looking for an alternative to Oracle or SQL Server. While it may not offer the enterprise-level horsepower of PostgreSQL, it still works well for many scenarios. This is likely why it maintains its status as the most popular open source database in the industry.

Recently, the folks behind MySQL released version 8.0 of the database. New features and functionality abound. Let’s take a high level overview of this new edition to see if there’s anything to help with your own software development projects.

The New Features in MySQL 8.0

The new version of MySQL added a whole host of enhanced SQL functionality. This includes support for window functions, common table expressions, as well as the NOWAIT and SKIP LOCKED statements. Most notably, window functions provide the ability to perform analytics on stored data; this is long-awaited feature as SQL Server added it in 2003.

They also added support for descending indexes which provides a performance improvement, especially when working with large datasets. The new GROUPING() function lets developers build datasets that distinguish super-aggregate rows from the results of GROUP BY queries. Both of these features were highly requested among the MySQL user community.

Boosting overall application performance becomes essential when using MySQL. Version 8.0 gives developers a new syntax for including optimizer hints. You simply place them using something similar to inline comment blocks after a SELECT, INSERT, UPDATE, REPLACE, or DELETE statement.

Additionally, the new version adds optimizer hints for a variety of INDEX and JOIN statements.  Now developers can control index merge behavior for each individual query or even the table order when performing a join. The MySQL development team feels the new optimizer hint syntax makes it easier to use while also boosting code readability.

MySQL adds Improved JSON Support

The JSON format is essential for web applications that rely on transferring objects expressed as a data structure. MySQL 8.0 improves its support for JSON in a myriad of ways. First, it adds extended syntax for ranges when using JSON path expressions.

The new JSON table functions lets you use regular SQL statements when working with JSON data. This is a boon for developers especially skilled in writing SQL queries. It essentially creates a relational view of JSON data.

Other new JSON features in MySQL 8.0 include aggregation and merge functions. A boost in sorting performance and the ability to perform partial updates are also welcome. The former helps to optimize large applications while the latter makes replication processes faster.

Other New MySQL 8.0 Features

Another significant new feature in MySQL 8.0 is support for GIS, including the Spatial Reference System. This lets applications using the database to easily calculate global distances given a LAT and LONG. The database now supports bitwise operations on binary data types, making the processing of IPV6 addresses easier.

The MySQL query optimizer also gets some improvements beyond the new hint syntax. Histograms and better handing of data buffering help engineers boost overall app and database performance. Finally, the database boasts improved reliability, availability, and reporting with an eye towards being used at companies following DevOps.

In short, MySQL 8.0 adds a host of new features making the database more attractive to organizations with high-demand applications. The improved DevOps support is also welcome. For more detailed information on MySQL 8.0, simply click on this link.

Thanks for reading this edition of the Betica Blog. Keep returning for additional insights on the software development world.

PostgreSQL is a Hot Database Choice yet Again

It seems the venerable PostgreSQL database is garnering a new wave of buzz across the IT industry. Maybe our series of articles on Postgres earlier this year helped contribute to this newfound hipness? Probably not, but modern tech organizations hoping for a cheaper alternative to Oracle while still getting support for NoSQL consider it to be a worthy option.

Let’s look more closely at some of the reasons why PostgreSQL remains one of the hottest databases – relational or not – on the market. Perhaps it makes sense for your team’s next data-centric web or desktop application?

Postgres is actually growing in Popularity!?

PostgreSQL’s increase in popularity caught the attention of InfoWorld magazine, who recently talked about the database’s hot factor earlier this month. In fact, Postgres now ranks as the 4th most popular database in the industry, according to a study by DB-Engines. Not surprisingly, the only three DBs ranked higher are Oracle, MySQL, and SQL Server.

The reasons for the growing popularity of Postgres – especially with younger developers – are numerous. InfoWorld’s writer, Matt Asay notes the improved performance brought by the support for JSON included in PostgreSQL 9.2 and boosted in version 9.4. Another important reason involves programmers growing tired of trying to fit even hipper NoSQL options like MongoDB into solution where a relational database makes the most sense.

Ultimately, in a situation when an old-school DB works best, PostgreSQL’s open source nature is simply more cost effective than Oracle or SQL Server. In fact, Postgres first earned its mojo as a cheaper alternative to Oracle. Still, could this old school database scale fast enough for use in modern web applications?

PostgreSQL and its newfound Scalability

The ubiquitous nature of social networks like Facebook and Twitter puts the onus on modern web applications to be extremely scalable. Most RDBMS options generally provide poor scalability, as did Postgres for most of its existence. Asay notes the introduction of Citus, an extension for PostgreSQL, provides a level of scalability rivaling many of the popular NoSQL databases.

Citus supports Postgres instances across multiple nodes, while providing a distributed model for transactions and SQL queries. These features give this veteran relational database the parallelism required for a massively scalable application able to compete in today’s market place. Take that, Cassandra.

While Citus is available as an open source extension, the company that developed it also offers a commercial version with full support. This is a similar model that EnterpriseDB followed with PostgreSQL itself. Citus provides a great option for shops working with Postgres for development and them implementing Citus for extra scalability before going live.

It also lets companies take advantage of their in-house talent’s database skills without spending on training in the latest NoSQL database options. These bonuses are arguably behind the still growing popularity of PostgreSQL.  Elijah Zupancic‏, the Director of Solutions Engineering for Joyent comments on some of the other core reasons.

“From a developer perspective, it is a pleasure to use. The documentation is wonderful, the data types reflect the types developers work with, and there is little surprising.”

Keep coming back to the Betica Blog for additional insights on software development, testing, and occasionally, databases.

PgBouncer and Pgpool – Essential for Scaling PostgreSQL

Companies looking at PostgreSQL as a production database alternative to Oracle typically want their web applications to scale quickly as those expensive commercial alternatives. The competitive modern business landscape – most notably the fickle online customer — demands high performance, and will look elsewhere when encountering a slow eCommerce site. Ensuring high scalability when using Postgres is a must!

Thankfully, the robust open source community around PostgreSQL developed two tools aimed to boosting the database’s performance on high traffic websites. Called PgBouncer and Pgpool, they need to be considered as part of any Postgres implementation. Let’s take a closer look.

A Lightweight Connection Pooler for PostgreSQL

PgBouncer serves as a connection pooler for Postgres. It is a lightweight tool known for a small footprint with minimal overhead. The program simply caches connections to different database servers. Its ability to manage a pool of database connections limits the number of actual connections to each Postgres instance; boosting the overall performance during a high traffic scenario.

Focusing specifically on connection pooling is one of the reasons PgBouncer is able to keep its small footprint. As such, it is not usable for other vital web database functionality, such as load balancing or replication. Pgpool is an option for handling those functions in addition to connection pooling.

Postgres Load Balancing, Replication, and more

Pgpool – now in its second major iteration, so Pgpool-II – is more of a Swiss Army knife of Postgres middleware functionality. In addition to connection pooling, it also performs load balancing, parallel query processing, replication, and failover handling – vital features for any scalable web application database. All client access to a PostgreSQL instance (or multiple instances) is managed through the Pgpool middleware.

Replication and load balancing are essential for any web application receiving high volumes of traffic. Load balancing lets Pgpool-II seamlessly distribute SELECT statements across multiple replicated database servers; improving overall system throughput as a result. As middleware, it mimics the API of Postgres, so a client application connects with Pgpool-II in a similar manner as any PostgreSQL server.

PgBouncer or Pgpool – what Option works Best?

Many system architects and/or database administrators probably wonder whether PgBouncer or Pgpool makes the most sense in their database application. Maybe using both is the best solution? The following database administration blog looks at this question while providing an overview of both tools.

PgBouncer by itself is the wisest choice if connection pooling is the only major need. Maybe a separate load balancer is already in place? This rule also applies to applications hosted in resource-constrained environments where quick database calls with low throughput happen frequently.

Pgpool naturally makes sense if you need its additional features, like replication and load balancing, combined with connection pooling functionality. Both tools work well together in applications where many database connections are expected and PgBouncer’s lower overhead in regards to pooling helps the overall performance.

Whether choosing Pgpool, PgBouncer, or a combination of the two, there’s no denying both tools remain valuable options for any high traffic PostgreSQL application.

Keep returning to the Betica Blog for additional information, advice, and insights from the rich world of software development. Thanks for reading!