Five strategies for WAN data acceleration

Flash storage, hyperconverged infrastructure, Linux containers and orchestration tools, bring your own device (BYOD), and software-defined networking are the top five strategies for data center operations, according to Bill Kleyman, director of technology solutions at EPAM Systems. However, one key issue was overlooked in his article on the topic for Data Center Frontiers: wide area network (WAN) data acceleration.

All data centers will be affected by data and network latency and packet loss, so consider the top five strategies for wide area network (WAN) data acceleration -- perhaps in addition to some of the strategies Kleyman proposed, such as software-defined networking. This will ensure that your data is secure and convenient, and reduce the adverse effects of latency and packet loss. This is critical when companies need to move large amounts of data quickly.

[[249624]]

Software Defined Networking

Kleyman writes in his article about software-defined networking (SDN): “As new applications require cloud delivery or deployment, the need to move data into the cloud is increasing. This requires a new approach to managing data through a layer of software on top of existing network hardware to drive innovation. Industry analyst firm Gartner predicts that by 2019, 70% of existing storage array products will also be available as ‘software-only’ versions. And by 2020, 70% to 80% of unstructured data will be stored on low-cost storage devices, managed by software-defined storage environments.”

Software Defined Networking (SDN) can do a lot for your data center. Imagine being able to recreate your entire network environment to mirror your existing infrastructure. The difference? Everything is network virtualized and completely isolated. Software Defined Networking (SDN) can help establish connections between applications, services, virtual machines, and many other workloads. In fact, managers can completely test their environment, disaster recovery (DR) plans, or high availability methods from a secure and isolated configuration. There, enterprises will focus on features such as network function virtualization (NFV), WAN traffic control, load balancing, security, WAN-OP, and even SD-WAN.

Precise selection

Kleyman selected key topics driving performance and cost savings within the data center, from flash storage to hyperconverged infrastructure deployed at the edge to the latest advances in SDN. However, with the push for cloud, hybrid cloud, edge computing, and more, few people are considering connecting these technologies together, yet these technologies are high on everyone's agenda.

However, while Kleyman looked internally at the top five data center strategies that will drive performance and efficiency, he did not consider performance inhibitors outside the data center. This is unusual because the data center is not an island. Data and access are the lifeblood of the business. People and organizations no longer just generate or consume data within the same data center.

The efficient flow of data in and out of the data center is essential to the effectiveness of any data-driven enterprise. Such is the criticality of data movement, everyone involved in the smooth and efficient operation of data must view data transfer holistically, not just in their sphere of influence.

Consider latency

In an article he published titled “Understanding the World of Edge Computing and IoT Devices,” Kleyman considers the impact of latency and packet loss on data transmission from remote IoT devices and the problems that can result.

Traditionally, organizations have tried to use WAN optimization to improve WAN throughput. WAN optimization is a bit of a misnomer because it doesn't address latency or packet loss. In reality, it's about optimizing data and applications to give the impression of improved performance. The solution is to add more computing power to the edge, closer to the Internet of Things (IoT) device or application, while providing some computing power for more time-sensitive applications while still transferring data back to the data center.

Data security

He also brought up another key point that touches on many issues, which is the security of data transmission over the network and wide area network. When using WAN optimization technologies that rely on deduplication to provide "WAN performance", using data encryption is a major problem because deduplication cannot be performed. So what about SD-WAN? Can SD-WAN solve the latency problem?

SD-WAN is a huge technological leap forward for small and medium WAN bandwidth applications, with its ability to bring together disparate WAN connections under a single software-managed WAN. However, it does not solve latency and packet loss issues, and any performance gains are typically limited by built-in deduplication technology. These issues can be mitigated to improve performance by adding a WAN data acceleration layer, which can move large amounts of data faster and more efficiently than SD-WAN.

WAN Data Acceleration

So WAN data acceleration is an approach that places itself outside of the traditional approaches and technologies of WAN optimization and SD-WAN. It takes a completely different approach to mitigate latency and packet loss issues. And the fact is that speed is not fast enough, which is why it is important to control latency and improve WAN performance over long distances.

While no one can fix latency, its impact can be managed by using TCP/IP parallelization techniques and artificial intelligence (AI) to control the flow of data over the WAN. Typically, organizations can see 95% WAN utilization as the effects of latency are mitigated. Another benefit of not using compression or deduplication is that WAN data acceleration will accelerate all data in exactly the same way solutions like WANrockIT and PORTrockIT do.

An IT person said that he had communicated with a marketing director of an IT company, who believed that IBM and other large IT vendors would solve the latency and packet loss problems. However, this was not the case. These companies also needed to deploy WAN data acceleration solutions. In fact, IBM uses its own WAN data acceleration solution in its products.

Still, this is a very common comment heard from many companies that are coming up with a completely different approach to this trend. In fact, one should ask these large OEMs why similar new approaches or other solutions have not been developed.

Five acceleration strategies

To summarize, here are five strategies for WAN data acceleration:

• There is no one solution that solves all problems, and the same is true for WAN optimization and WAN data acceleration. Therefore, consider developing different levels of potential solutions.
• For Office-type file transactions over slow WAN connections, WAN optimization should be the first port of call for enterprises.
• If the files or data you are transferring over the WAN are encrypted, WAN optimization will not add any benefit. Instead, consider deploying WAN data acceleration, as it can improve performance even during periods of low WAN bandwidth and high latency.
• Use WAN data acceleration when the WAN is working at its best, with constant data flow, to match or exceed the performance of the WAN.
• Keep in mind that WAN data acceleration efficiency increases exponentially with WAN bandwidth and latency. With a 10Gb connection and 30ms latency, customers can experience up to 200x performance.

Improve performance

One of the benefits of using WAN data acceleration is that it treats all types of data the same - it doesn't matter what the data is. This feature allows it to reach other areas, such as a storage area network (SAN). By separating the data from the protocol, organizations can transfer data between SAN devices across thousands of miles.

For example, one customer connected two virtual tape libraries over 2,860 miles over WAN bandwidth, and the performance was improved by 95 times the original performance. Therefore, even in disaster recovery scenarios, it should be implemented as a strategy to shorten recovery time and improve service continuity.

Traditionally, data centers are often located within the same outage zone, but with WAN data acceleration, processing can be done in two locations thousands of miles apart. This would allow data centers to continue operating in the event of a natural or man-made disaster, for example. Among other things, this capability can save a business from financial and reputational damage while creating a competitive advantage. WAN data acceleration is therefore a flexible solution with artificial intelligence and machine learning that can make a big difference.