SaaSification is not only a tech thing. Entire industries are moving to more specialized services provided as applications and APIs. While finance is the poster-child for this move, SaaSification crosses every industry from consumer-facing to B2B. The growing move toward SaaSification is a topic I covered in depth in a recent InfoWorld column. I looked at how companies can SaaSify themselves to become more valuable and finally how IT departments can prepare for these changes. In this blog, I explain how a distributed SQL database, MariaDB Xpand, accelerates SaaSification.
SaaSification Breeds More Data, Fast
When an industry moves from each individual company recreating the same service to a few competitive vendors, the data is more concentrated and intense. Think of vendors like Expensify or Salesforce. When an individual company ran its own customer relationship management (CRM) or expense reporting tool, there was less data to manage at each instance of the CRM tool. When they moved to something like Expensify, one vendor had to manage much more data.
Expensify started out with a client-server database architecture with a proprietary replication tool and it was crushed multiple times. However, it was not just their database architecture choice, they also had a spike in user traffic around April 15th, the traditional date for US income tax reporting. Expensify solved many of its problems by developing its own distributed database with a SQL backend. While that worked for Expensify, most firms do not want to, nor do they have the resources to develop and maintain their own database system.
SaaS companies need to handle massive-sized databases to store their ever-growing data, but also a tremendous number of reads – sometimes in the millions. Meanwhile, they also need to handle massive numbers of writes. Traditional read replicas usually do not cut it. For financial firms, SQL databases are the norm because data integrity is essential, and SQL is the lingua franca of analysts. Finally, SaaS companies – especially finance – need to enable customers to run reports. A purely transactional database is not going to cut it. There is no time to wait on data to titrate to an analytical database just to be able to run a report.
One thing a SaaS company cannot do is lose its customers’ trust. The service must be reliable and, above all, cannot lose data. This means not only high availability but constant online backups and geo-redundancy. The system must survive the big one, even a bomb cyclone in the fall or early winter, or a massive heatwave in the summer.
These problems are bubbling up to everyone. Data volumes continue to grow at an increased velocity. Meanwhile, business requires “real-time” information that requires a greater concentration of data. Any outage, even internal, affects the bottom line in a competitive market where it cannot be afforded. Disparate systems with stove-pipe databases are no longer enough. Even internal systems need SaaS reliability, for example, to handle scheduling flight crews in response to weather events. This change throughout the economy means that we are all SaaS companies, eventually.
MariaDB Xpand Distributed SQL Database for SaaSification
Xpand answers this call. It supports horizontal scalability and splitting of the database across a large cluster. This architecture enables ever-increasing database sizes. Applications see a single database, unburdened by data distribution details, while Xpand is distributing reads and writes evenly across the cluster. Xpand has been proven to scale in production in the millions of queries per second. We’ve benched it in the tens of millions (stay tuned for more on that in a later blog). When it comes to long-running queries, Xpand uses a columnar index feature that allows developers to run the types of reports and analytics that their users need. Data can be replicated in parallel to multiple regions, and online backups can also be done in parallel while maintaining transactional integrity. Finally, an IT department doesn’t have to become distributed database experts because Xpand is also available as a fully managed cloud database service through MariaDB SkySQL.
Xpand is exactly the kind of technology a SaaS company or a SaaSifying company needs to manage ever more data, spikey user traffic, and analytics on demand, with global reliability. Distributed SQL is the kind of technology that financial firms, among others, are relying on in order to meet new demand and enable new business models. Xpand’s unique feature set combining MySQL/MariaDB compatibility and transactional capabilities with operational analytics make it unique among distributed databases.
Try MariaDB Xpand
Get started with MariaDB Xpand as a single node docker image or try its full capabilities in the cloud on SkySQL.