m-Power now offers in-memory data processing

August 2015 · Oak Brook, IL

Today, software developer mrc announces the release of in-memory data processing within m-Power. m-Power is a web application development platform that businesses use across their organizations. It creates web applications such as report-writing, Business Intelligence, executive dashboards, customer portals, and mobile applications to name a few.

What is in-memory data processing?

In-memory data processing involves querying data stored in Random Access Memory (RAM), as opposed to data stored on physical disks. As the cost of RAM declines, the use of in-memory data for tasks like reporting and analytics has grown in popularity. Because memory offers faster response times than traditional database calls, the in-memory approach offers improved performance for certain application types.

What does this enhancement offer?

This enhancement brings in-memory data processing to m-Power. Users can now create applications over in-memory data sets, or over traditional database tables. These options give m-Power users the best of both worlds: Some applications are better suited for in-memory data, while others work best with traditional database data. For example, the in-memory approach works best when a business needs to:

• Perform interactive analysis over large data sets. When working with large data sets, interactive applications (like interactive reports or pivot tables) will benefit from in-memory data. Over a traditional database, response times can range from seconds to minutes, depending on database speed and data volumes. These slow response times can make the report nearly unusable, as users must wait every time they make a selection. Running this same report over in-memory data can reduce response times dramatically, and deliver faster reporting.
• Relieve the burden on the database. A database is responsible for a company's transactional systems. Performing analytical queries on the database can slow it down, especially if it's processing other tasks at the same time. If these analytical queries hurt the database's performance, a business can perform analytics over in-memory data to relieve the burden.

However, while in-memory data is well suited for the examples above, it's not the best choice for every situation. For example, traditional database applications work best when a business needs to:

• Create applications over real-time data. While in-memory analytics offers fast queries, it doesn't provide real-time data. The data must be first taken from the database, and brought into an in-memory data set. For applications that require real-time data, running them over live database data is the best option.
• Run reports over small data sets, or a fast database. Many businesses aren't working with large data volumes, or their database is fast enough to handle the workload. As such, in-memory data would only deliver a marginal performance increase. In these cases, connecting directly to the database and avoiding the extra step of moving data to memory is the best option.

"In-memory analytics is a growing trend, and offers many advantages," says Tyler Wassell, mrc's Director of Development. "But, it's not the best choice for every situation. With this enhancement, m-Power users get the freedom to choose the best approach for their needs."

###