7 Advantages to Running Hadoop on Isilon

The rate at which new data is being generated daily is staggering and can be overwhelming for many; every single person on the planet is currently generating 1.7MB of data per minute, which will equate to 44 trillion GB of stored data by 2020! But not all data is created equal…

This explosion in data growth does not necessarily result in uniform loads on infrastructure, so when enterprises are selecting a storage solution, they must be able to recognize and adapt to the different growth rates required.

Dell EMC’s Isilon Scale-Out Storage Solutions for Hadoop combines a powerful yet simple and highly efficient storage platform with native Hadoop integration that allows enterprises to accelerate analytics, gain new flexibility and avoid the costly activity of deploying a separate Hadoop infrastructure.

Let’s take a closer look at some of the key advantages of running Hadoop on Isilon:

1. Scale-Out Storage Platform

Multiple applications and workflows within an organization can benefit from scale-out storage by no longer requiring DAS-based Hadoop clusters for their own purposes. Maximum flexibility is provided by Isilon with one large bucket of storage being used for a variety of simultaneous uses, including multiple simultaneous Hadoop distributions.

2. No Single Point of Failure

Traditional NameNode configurations present the possibility of a single-point of failure due to the loss of a non-clustered namenodes. With Isilon, we have enabled every node in the cluster to act as both a namenode and datanode, thus resulting in a highly resilient infrastructure.

3. End to End Data Protection

Datasets are protected by Isilon’s highly-available architecture, and snapshot and replication capacities, thus offering unprecedented data protection and security. With this in mind, unwieldy copy data backups are not necessary when leveraging Isilon for datasets of any size.

Furthermore, the built-in data and hardware protection capabilities of Isilon created two additional benefits that are absent in traditional Hadoop, these include:

  • Namenode redundancy without having to create multiple namenode instances.
  • And dataset data protection against node loss and corruption.

4. Industry-Leading Storage Efficiency

There are a dizzying number of storage solutions out there, and with space being a premium (whether it’s digital or physical), choosing the right storage solution must be considered a priority. With Hadoop built natively into Isilon, enterprises can be rewarded with less than 80% storage utilization resulting in industry-leading storage efficiency and lower costs.

Isilon’s operating system dedicates a smaller portion of the overall capacity to redundancy which offsets much of the cost benefit of commodity storage, and in some cases, Isilon comes in as less expensive than commodity.

5. Independent Scalability

In today’s data driven economy, storage needs often outstrip the demand for compute resources, but adding more servers to “fix” these storage problems often results in underutilized CPU resources. Therefore enterprises need to understand which workloads are storage intensive versus compute intensive.

By running Hadoop on Isilon, enterprises have the ability to add compute and storage nodes separately, thus providing customized solutions that can scale with their business easily and cost-effectively.

6. Separation of Compute Nodes from Data Nodes

In addition to adding compute and storage separately, compute nodes are separated from data nodes. This unique setup feature of Isilon allows each node to boost performance and expand the cluster’s capacity. For Hadoop analytics, the Isilon scale-out distributed architecture minimizes bottlenecks, rapidly serves big data, and optimizes performance for MapReduce jobs.

7. Multi-Protocol

Isilon’s revolutionary infrastructure permits data to be stored in its raw format. This means that it can easily be exported via the wide variety of industry-standard protocols supported by Isilon without copying the data, including NFS, CIFS, FTP, HTTP, and HDFS, to name a few. This unique capability enables data to be stored on an enterprise storage platform with existing workflows and standard protocols.

See how All Flash solutions have provided lifesaving discoveries:

http://www.youtube.com/watch?v=gZsOj59UbM0

As the first scale-out NAS platform purpose-built for Hadoop, Isilon has the capability to provide a powerful analytics platform with a huge centralized storage. The integration of Hadoop also reduces the effort associated with disk failures, eliminates the typical 3x data replication for traditional configurations and solves the need for moving and staging data to make it accessible for Hadoop.

For many customers, running Hadoop on Isilon represents an optimal balance of scalability, availability, and performance while reducing the operational expenses associated with the maintenance of Hadoop clusters. And while there is no one-size fits all, Dell EMC’s Isilon solutions seem to tick many more boxes than our competitors thus equipping your business with the tools to handle data growth and scale accordingly and affordably. Visit our website here to see all of our Isilon hardware options.

About the Author: Dell Technologies