Summary
In the fast-paced world of business companies often need scalable, reliable, high-availability database solutions. Any enterprise application developer, using Microsoft SQL Server as the database solution faces many challenges in terms of increasing application availability, which often means using multiple database servers for data transaction processing. SQL Server administrators face the challenge of monitoring status of multiple servers, often times one of which is the mater server and the others are backup servers. When the master server fails it's challenging to switch transaction processing to backup servers. Many of these problems get solved by Microsoft Enterprise-level products (clustering etc.), which is quite costly.
TDS team has come up with a cost effective solution, which makes it easier for Microsoft .NET developers to interface with multiple redundant Microsoft SQL Servers, providing essential clustering features, such as Load Balancing, Failover, Read/Write Affinity, Status monitoring.
The Solution
To balance server load, TDS created the system which distributes requests to different nodes within the virtual cluster, aiming to optimize system performance. This results in higher database availability and scalability - must have for any modern enterprise-level database solution. TDS DataFarm provides the companies with an application package, which is easy to install and operate.
TDS DataFarm provides the following functionality:
- Load-balancing SQL queries to a database server, providing efficient distribution of the workload. If one server cannot handle a request immediately, another server takes over.
- Fail-over capability of the TDS DataFarm allows for automatic switching over to a healthy server upon the failure or abnormal termination of the previously active server.
- Affinity makes it possible to fix the open connections between the specific user's transactions and a specific server when necessary.
- Read and Write balancing provides the ability to set the rules for Read, Write, Read-and-Write, Read-then-Write options.
- Partitioning capability provides efficient capacity utilization by sharing the data between the clusters nodes. The TDS DataFarm partitioning functionality makes it possible to send the user's request to a pre-defined 'partition' server, which contains the specific requested partition, where the necessary data is stored.
- TDS DataFarm is easy to manage via the Microsoft Management Console. The administrative user interface makes it easy to configure and monitor every node via a familiar tool set.
- Easy to install: the user is able to deploy DF from single host to all necessary hosts.
- Easy to use: easy to configure and interoperate with SQL databases. Intelligible and convenient GUI simplifies the work with DataFarm.
TDS had evaluated the technological needs of this project and put together a dedicated team of testers and developers with expertise best suited to perform the work in the most effective way. The team consisted of test engineers, 3 development engineers (2 Microsoft Certified Professionals), a test lead and a development lead. The product was developed by this team in 5 calendar months.
DataFarm Benefits
The TDS DataFarm technology was successfully integrated into the Iceportal TM web portal. The portal is reliably serving 2,000,000 read and on the order of 10,000 write transactions per day. The investment made in developing this technology has paid off by letting Iceportal use a more cost-effective combination of Microsoft software as well as maintaining top-notch reliability and data integrity.