Modernizing SQL Server Applications, Part 2: Your Top 3 Questions Answered
Here are three big questions that we find that many businesses are asking us about Tarantool these days and their answers.
Join the DZone community and get the full member experience.Join For Free
For Part 1 of this series, an overview of SQL Server and Tarantool, please use this link.
You can offload large workloads that slow down SQL Server and drive costs to the in-memory database Tarantool, a unique product that is fully ACID and that features a SQL:2016 implementation. This will allow you to downsize your enterprise servers and save on licensing costs. Tarantool also includes a complete Lua application server that will open up a world of functionality to your stored procedures.
Tarantool’s replication of SQL Server is part of Tarantool Enterprise and is enabled by its partners Continuent and Informatica. You can also drive a Tarantool database from your C# code using the open-source ProGaudi connector.
Here are three questions we find that many businesses are asking us these days:
How does Tarantool differ from Microsoft’s In-Memory OLTP (formerly known as Hekaton)?
To begin with, both Tarantool and In-Memory OLTP support stored procedures. However, not all T-SQL constructs are supported by In-Memory OLTP (here is the extended list of the unsupported constructs). In contrast, there are very few limitations to what can be programmed using Tarantool’s full, non-blocking implementation of Lua for stored procedures.
A further consideration in relation to In-Memory OLTP is that Tarantool can be replicated, whereas In-Memory OLTP tables cannot. And finally, keep in mind that Tarantool was designed from its first commit in 2009 to be a self-sufficient in-memory product, not retrofitted to accommodate an older technology.
What else does Tarantool let me accomplish that SQL Server and its product line do not?
Adding a versatile technology like Tarantool to your stack will allow you to rearchitect it, in addition to making it faster (it may end up facilitating a whole new way of thinking about how to design your implementation). You could, for example, completely remove the responsibility of authenticating users from SQL Server and let Tarantool assume the task. Of course, an additional advantage of Tarantool is that it allows you to avoid complete vendor lock-in.
Great — but is Tarantool hard to use?
Tarantool is easily installed with various Linux package managers, Docker, or Homebrew. You can also use WSL and there is an Azure image. As far as Lua goes, its syntax is one of the simplest out there and we find that people pick it up relatively quickly (don’t let its simplicity fool you, though; it is quite powerful).
Once you install Tarantool, it includes a tutorial that you can run with
tutorial(), which will get you up to speed on Lua as well as some of the features of Tarantool. You can also have a look at our Tarantool 101 series, which shows you how to program and run a simple Tarantool microservice on Digital Ocean (no additional backend language needed). Finally, our Enterprise support packages will allow you direct access to members of our core engineering team.
Tarantool is versatile in the ways it can improve your SQL Server stack, whether you want to offload a few expensive queries or whole business processes. Please get in touch with the Tarantool team here to tell us more about your specific project.
Opinions expressed by DZone contributors are their own.
Using OpenAI Embeddings Search With SingleStoreDB
Tactics and Strategies on Software Development: How To Reach Successful Software [Video]
Knowing and Valuing Apache Kafka’s ISR (In-Sync Replicas)
Database Integration Tests With Spring Boot and Testcontainers