DataCore Provides Performance for SQL Server Installations without Breaking the Bank (Part 2 of 3)
Introduction
In the previous blog we discussed that IT organizations need to maintain service level agreements (SLAs), and the challenges they face in order to deliver superior performance for their SQL Server environment. In most cases the performance problem is due to the underlying storage, but the current storage options are expensive, complicated, incomplete or disruptive. DataCore offers a comprehensive solution to overcome these concerns. In this blog we will focus on the performance aspect of DataCore.
Common approach
The common approach for storage has been to allocate all cores to compute except for one that is assigned to IO. As a consequence, the gap between compute power and IO has become a gaping chasm. The best analogy to explain this challenge was narrated by colleague, Augie Gonzalaez, Director of Product Marketing.
He said;
“Imagine a cashier line at a cinema. There are five cashier booths available but only one is manned. Half an hour before movie time, long lines develop and that one lonely cashier struggles to get everyone inside before the start of the feature. That is how the common approach handles the IO, in which IOs develop a long queue and applications slow down. Now imagine all five cashier stations manned; traffic flows through without slowdown. This is the DataCore approach.”
DataCore’s approach
DataCore™ Parallel I/O and high-speed ‘in-memory’ caching technologies speed up SQL Server I/O response both for reads and more importantly for writes using storage resources already in place. These breakthrough technologies solve the performance problem for the SQL Server environment by exploiting the multi-core environments to the maximum. This new way ensures that there is never a “wait state” for any of the cores and all cores work in parallel when there is IO demand. As a result, DataCore increases user productivity by dramatically reducing the SQL Server query times and speeding up jobs and logins. Unlike other storage options, this performance increase does not require investing in expensive hardware, or wasting time and effort fine tuning databases to meet SLAs (figure 1.)
Figure 1
Customer Example
PWS Distributors are one of the UK’s leading designers and suppliers of quality kitchen components and work surfaces to retailers and installers, offering over 8000 stock items either on the phone through Customer Services, or via their e-commerce site which accounts for more than 20% of business. The company thrives on providing tailored customer service but in the background major processing issues were restricting the growth of the business and orders. Problems first showed within the ordering cycle when new items were added to an existing order, or when orders were changed in the ERP (Enterprise Resource Planning) system.
With DataCore, they experience 14x performance increase. Their customer services efficiency has been bolstered, with an active working day saving of over 2% for every representative waiting for ERP system responses, and users no longer have to wait for the system to catch up with their input. Making modifications and additions to the system has improved significantly. Nightly invoice processing has reduced from 2 hours to 20 minutes. Daily management reporting on best-selling lines has been transformed from 30 minutes to an incredible 10 seconds. Other reports have also been impacted. Aggregated customer statistics queries’ running millions of SQL executions used to take over 24 hours, even in low trading conditions. With this solution, mining takes just an hour.
Click here to read the case study
Next blog
In the next blog, we will touch more on the availability aspect of DataCore.
