Everything You Need to Know About Scaling MySQL Part 8: Wrapping Up

We have come to the end of the line for our “Everything You Need to Know About Scaling MySQL” series and we hope you have come away with some valuable insight into how to optimize your e-commerce website. For those who may have missed an installment or two—or simply need a refresher—here is an overview of the key takeaways from each of the previous seven pieces:

Part 1: Websites are processing far more transactions than ever before. As such, legacy scale-up databases are now encountering problems with MySQL performance and, as a result, website availability and performance.

Part 2: MySQL can handle a large volume of traffic, but at some point it inevitably hits a wall and runs out of capacity—at that point, the site will no longer function optimally. Furthermore, because it conforms to a master/slave architecture with a single point of failure, if the master server goes down, the site goes with it.

Part 3: When a MySQL write-master is scaled to capacity, e-commerce sites have the choice of either replatforming (migrating from one database solution to another) or sharding (splitting up a database across multiple servers). Both options are expensive and complex, however, and e-commerce sites should consider breaking away from legacy database architecture.

Part 4: Facebook and Google have had their own difficulties with MySQL and poured immense resources into solving those problems. Most companies don’t have the same resources as these two tech giants, however, and must find other ways of combating MySQL performance issues… including abandoning MySQL for a NewSQL solution, as Google did.

Part 5: Regardless of how a company chooses to scale MySQL, it is a time-consuming and difficult process. As you scale you will have to commit more manpower to the project and it becomes more difficult to maintain ACID compliance as well.

Part 6: Amazon’s new MySQL-compatible database, Aurora, has no simple solution for scaling writes. Aurora was constructed using a single write master and read-slave architecture, meaning performance and latency issues seem inevitable.

Part 7: Ultimately, with a legacy scale-up relational database, e-commerce customers are the ones who suffer. They may experience sluggish performance or even a crash, which means they may take their business elsewhere.

There’s always more to learn about how to get the best possible performance from your e-commerce website. See our site for additional resources.