The Database Landscape
The database landscape is vast. For multiple decades, traditional SQL databases have been the database of choice. However, in the last decade they have struggled to keep pace with market demand. With new sources of data and larger data sizes, traditional SQL databases struggle to support millions of users, billions of data rows, and billions of reads/writes around the clock. Users also now expect to run analytics on bigger datasets and get real-time insights into their business in order to make agile decisions. These expectations have challenged the limits of traditional SQL databases, which have decades-old code, are not distributed, and were never designed for the scale-out architecture of the cloud.
A variety of new players have entered the database landscape to address these challenges. While some of these databases are strong solutions, many others don’t deliver on their promises. However, if you evaluate each player on fundamental technology, you can see the real strengths, and weaknesses, of each product. It then becomes a question of choosing the best solution for your particular problem.
Why Did Clustrix Come Up with its Own Database Landscape?
The influx of new players in the database market has lead to confusion. Analysts work hard to provide complete information, but all of the marketing hyperbole makes it challenging to fully understand the landscape. There are also several limited solutions that are not new databases and thus don’t possess the features and capabilities necessary to be a long-term, viable solution in the primary database market.
With this in mind, we decided to come up with a digestible landscape graphic that looks at each player’s fundamental technology choices. Our goal is to help companies find their own ideal database solution.
What About NoSQL?
We decided to exclude NoSQL database solutions in this landscape in order to simplify things. For certain kinds of workloads, NoSQL solutions casino are a good fit. Some new web applications have non-critical data that can be inconsistent and might even get lost. However, companies that require SQL features like ACID guarantees (e.g. e-commerce) should consider scalable SQL options. From what we see in the market, those companies that decide on NoSQL solutions feel the pains of lost data over time, and then build SQL features in their applications. We’ve seen a flurry of posts recently about companies that have lost their data because NoSQL databases don’t have ACID guarantees.
What’s Happening with NewSQL?
Traditional SQL databases (such as MySQL) provide good performance for both transactions and analytics, but don’t scale beyond a single machine. They’re being replaced by four distinct categories of distributed SQL databases that scale well and can excel at practically any workload—removing any remaining doubts about SQL.
The first two categories of NewSQL databases are the in-memory row and column stores. These databases are good at low-latency (<10 ms) transactions and fast OLAP, respectively. However, because they focus on only one of the two workload types, neither can do fast updates and analytics on the same table.
The third category of NewSQL databases is the SQL columnar stores. This has been established in offline analytics and data warehousing and is beginning to see widespread use.
Clustrix is the only database solution that fulfills the last category of NewSQL databases—shared nothing row store. Clustrix is able to run mixed workloads similar to older databases such as MySQL and MS SQL Server. With Clustrix, you can scale transactions and run fast real-time analytics by simply adding more nodes. And as of May 2013, Clustrix 5.0 is available on the AWS Marketplace. The database chooses the combination of row stores that are ideal for transactional speed and couples them with massively parallel processing (MPP) akin to the columnar stores for fast analytics. Looking at the needs of today’s market, Clustrix brings together a breadth of features that makes it an ideal primary SQL database for companies in various industries such as SaaS, e-commerce, gaming, healthcare tech, and consumer web services.