Make It Simple

The NuoDB emergent database is a pioneering, SQL-compliant database management system designed from the ground up for resilient, continuous operation across virtual and cloud computing environments. Here are some common uses for the NuoDB solution:

Building new web-scale applications

One day you are a guy working out of your garage, serving up a web app to 200 users, a lot of them friends and family. Jump ahead 365 days and you’re a guy whose app was named of best of class on Gizmodo. Now you’ve got 2 million users and that number is growing by tens of thousands every week. It’s a good problem to have. It’s good you saw it coming and built your database on NuoDB.

You knew you’d need to scale out fast; you need to have the ability to run multiple instances of the application in the cloud and the transactional database that supports your app better move with you easily. This is the first use case NuoDB was designed for: building brand new web-scale applications that could be tiny today and Facebook tomorrow.

A Customer Use Case: Our customer is a well-known application developer in Israel. He has been cited on Gizmodo and ZDnet for his innovative web apps. His current project is the creation of an app that consumers can use to rate photos on dating sites.

The Issue: He wanted to continue to use a relational database and specifically SQL syntax with which he is very familiar. But he also needed high availability and specifically the ability to scale on Amazon EC2.

What He Tried: He had typically used SQL Server and MySQL in previous projects but knew they would not scale as easily as he needed them to. He considered a switch to NoSQL, specifically Amazon DynamoDB, MongoDB and Cassandra. After some experimentation, he came to the decision that he was unwilling to give up the combination of high availability, SQL and ACID guarantees.

Results: He easily set up NuoDB locally on his computer and tested it in the cloud. Then he initiated a node failure on NuoDB to see how that would impact his application. Everything continued to work like nothing happened. Good start.

He was able to query the database using SQL language in which he’s an expert. As an individual developer, the ability to start and finish the app development project in a timely manner was also critical. With NuoDB, since there was no need to learn a new technology, or learn how to configure his application to support it, he was able to get his project completed on schedule.

Finally some developers don't care about scaling when their applications are first released; they’ll deal with that problem when the time comes.

Migrating an existing SQL database or offering SaaS

You wrote your database based on a traditional SQL platform and it’s huge. It takes a team of 10 SQL developers and database administrators to keep it running smoothly. They’ve sharded, cached, replicated, clustered it and done every database maneuver in the book to make it scale as fast as your application. It costs a bundle to keep it running. Bigger machines. More DBAs. More support costs.

It will be a painful experience to migrate that existing SQL database to a private cloud. But your CIO has done the math and knows the economic benefits of app virtualization and cloud deployment are so significant that the cost savings will outweigh the pain. The deployment flexibility, on-demand capacity, better uptime and higher performance are all enticements too. The CIO’s strategy is to build fewer, more flexible, commoditized datacenters and a private cloud, he says, “start to migrate. “ Here’s what a real customer did when facing that scenario:

A Customer Use Case: Our customer is part of a global SaaS (software as a service) company. They work with nearly 1000 financial institutions ensuring these bank customers are in full compliance to an array of national and local regulations.

The Issue: The customer’s current RDBMS is Oracle 10g. As the popularity of their SaaS has grown, Oracle is becoming exponentially more expensive to scale, support and maintain. In addition, the customer is uncomfortable having all their “eggs” in one very expensive vendor basket. In addition their need to scale has been labor intensive and takes too long.

What They Tried: The company has moved some portions of their application to Microsoft SQL Server and invested heavily in optimizing their Oracle implementation.

Results: This SaaS vendor is running a proof of concept with NuoDB to validate that scalability and performance meet their needs. Installation of NuoDB took minutes and the POC is ongoing smoothly. We’ll keep you posted!

Another NuoDB customer is also a leading global software company who wants to offer customers a SaaS version of their very complex application. The application contains millions of lines of code developed over the past 25+ years. Their current database is also Oracle and their challenges are the same:

• Expense of the Oracle database

• Difficulty in scaling

• Avoiding vendor lock

• Creating an easier-to-use, more differentiated SaaS offering.

Mobile and web-scale applications

According to research done by the Pew Institute, half of adult smartphone owners have downloaded at least one mobile application. The Apple App Store offers more than 500,000 apps, which in aggregate have been downloaded well over 5 billion times! All these applications are associated with a database. If a mobile application goes viral like Angry Birds, that database better scale very quickly and very easily.

A Customer Use Case: That is just the scenario facing our US-based customer who is a hot startup blazing a trail for their pay-by-smartphone application. The app ties into a database that contains customers’ credit cards, purchases, etc. The company’s mobile app is supported at 100 retail outlets today and growth is predicted to rapidly reach 5,000 outlets in the immediate future.

The Issue: This customer’s primary challenge is economic. They are presently hosted on a public cloud. The greatest use of their application occurs on weekends when tens of thousands of read/write transactions are being made against their database simultaneously. During the weekends they need to scale out. When Mondays and Wednesdays roll around, they are typically much slower days for credit charges. To save on the public cloud meter, the customer wants to scale back down on slower days. This frequent need to scale out and in needs to require very little effort.

What They Tried: Initially they built their database on PostgreSQL. They had scaling issues from the start so they quickly switched to MySQL. They have already determined this will not give them the elasticity they need in the future.

Results: “Because we have written thousands of lines of code in SQL, and we understand it so well, we have to have something that is standardized but also scales. Other solutions we have tried offer some degree of SQL compliance but they lull you into a lot of customization that NuoDB does not require. Also we need to be able to handle thousands of reads AND writes simultaneously when charges are all happening on a weekend. Many reads and one write at a time just don’t work for our application. NuoDB lets us do both at scale.”

Big Data

How much, how fast, and what types of data can your DBMS handle?  Volume, velocity and variety are the three dimensions that are most often used to describe Big Data. While there are many Big Data solutions that tackle Big Data problems, none of them are able to address the bigger challenge of providing SQL and ACID transactions. NuoDB is 100% SQL, 100% ACID and its emergent architecture is especially well suited to take on the challenges posed by the three V’s of big data and deliver big on the fourth “V” --- value.

A Customer Use Case: Like many others, this NuoDB beta customer is just starting to investigate Big Data. They are a leading nuclear power systems provider. For now they are researching ways to move two of their enormous database tables of 100 million records each. They have a federated file system and several remote locations. Their central database tracks file location and manages the lifecycle of each file as well as registering metadata.

The Problem: "Joins are too difficult. Oracle is too expensive. Queries have to scale quickly and right now they don’t.”

What they tried: This customer is currently using a combination of Oracle, PostgreSQL and MySQL to meet their needs. They are investigating NuoDB, NoSQL and HBase for resolution of their pain.

Results: NuoDB is going to win. Why? Our distributed transaction engines are designed to take on the real-time data processing needs of an OLTP system and simultaneously take on responsibility for backing your real-time operational intelligence needs. NuoDB also scales easily to accommodate volumes of data through integrated HDFS support for high-speed I/O at the storage layer. Plus all this without having your developer to learn a new query language.