POWER > Business Strategy > Competitive Advantage

A Deeper Look at POWER8 CAPI and Data Engine for NoSQL

NoSQL Benefits

In addition to the application acceleration solutions, CAPI can be applied to another current IT issue—available memory. CAPI can also be used as a base for flash memory expansion, as is the case for the IBM Data Engine for NoSQL.

The use of NoSQL has exploded in recent years as new customer-facing applications require unprecedented response time and scale to meet mobile user expectations. Typical NoSQL implementations run entirely in memory or rely heavily on memory as a cache; therefore, they become expensive and hard to scale. The latency of traditional I/O attached storage doesn’t always meet the application response time requirements, and a new solution is needed to support the growth in applications based on NoSQL data repositories.

Redis Labs reduces the nodes required by up to 24x and the deployment cost of a 12 TB database is 1/3 the cost of traditional deployment

The IBM Data Engine for NoSQL uses the CAPI platform to create a new tier of memory by attaching up to 40 TB of auxiliary flash memory to the processor without the latency issues of traditional I/O storage. This creates a new level of cost-performance between expensive DRAM and higher latency I/O attached storage. The latency falls within the acceptable limit of applications especially when data is accessed over the network as in most NoSQL solutions.

Because a significant amount of flash can be attached to a single node, most solutions will need far fewer processing nodes to deliver larger NoSQL solutions. This, in turn, reduces the deployment and operational cost for delivering the customer solution.

In contrast, NoSQL solutions that use flash memory as solid-state drives on the I/O bus (not CAPI attached) perform better than a spinning disk; however, they still are not able to meet the NoSQL latency times required because of I/O overheads compared to DRAM. There are also flash DIMMs, but they have size, resiliency and other performance limitations.

The Data Engine for NoSQL doesn’t lock you into a specific flash technology. By leveraging a standard connection to the IBM FlashSystem* 840, the flash size and performance can increase as new models of the flash array are delivered. By providing direct application access to flash, applications can have the flexibility to manage the data in ways that provide direct application benefit. Enterprise data centers, managed service providers and Internet service providers can all benefit from applying this new technology in the NoSQL application space. Exploiting the hardware and software built into the flagship POWER8 open architecture means that clients no longer must choose between “big” or “fast” for their solutions.

Redis Labs joined with IBM to be the first NoSQL provider to deliver using this technology. The company saw a unique opportunity to leverage flash memory attached to an open POWER8 processor through CAPI. By using a combination of DRAM and CAPI-attached flash memory, IBM and Redis Labs provide a solution that delivers a significant reduction in both deployment and operational costs while making it possible to power faster, larger and more scalable data repositories. For example, the deployment cost of a 12 TB database is one-third the cost of traditional deployment. Reducing the nodes required for the solution by up to 24 times results in a dramatic reduction in the total cost of operation (TCO) for networking, floor space, energy, cooling and operations overhead. By providing an adjustable mix of flash and DRAM in the solution, the client has the capability to make performance/cost optimization decisions. While Redis is the first to exploit this solution, others are following.

Brad L. Brech is a Distinguished Engineer at IBM.

Bruce Wile is a Distinguished Engineer at IBM.

comments powered by Disqus



2019 Solutions Edition

A Comprehensive Online Buyer's Guide to Solutions, Services and Education.


5 IBM i Architectural Principles Drive Business Innovation


How the API economy is changing business

IBM Systems Magazine Subscribe Box Read Now Link Subscribe Now Link iPad App Google Play Store