Get fresh updates from Hortonworks by email

Once a month, receive latest insights, trends, analytics information and knowledge of Big Data.


Sign up for the Developers Newsletter

Once a month, receive latest insights, trends, analytics information and knowledge of Big Data.


Get Started


Ready to Get Started?

Download sandbox

How can we help you?

* I understand I can unsubscribe at any time. I also acknowledge the additional information found in Hortonworks Privacy Policy.
closeClose button
May 13, 2015
prev slideNext slide

HP and Hortonworks – Driving Hadoop Forward

Since the launch of Hortonworks Data Platform (HDP) three years ago, we have seen first hand how Enterprises are embracing Apache Hadoop to enable their modern data architecture’s and power new analytics applications. Hadoop is helping organizations transform their business by providing them with a pervasive, enterprise ready data platform to meet their big data challenges.

Apache Hadoop’s ability to process any data (i.e., clickstream, web and social, IoT, etc.) allows an Enterprise to derive insights in ways that were previously either technologically or economically not possible. As they start to capture and process all these new and traditional data sources, Enterprises are looking into infrastructure that will address their big data requirements and needs, such as scale and storage density.

This opportunity to provide more value to the Enterprise is why we’re very excited to be part of HP announcement of their new HP Apollo Big Data Server family. This new line of servers takes an innovative approach on how Enterprises can leverage infrastructure to build out their modern data architecture.

In this guest blog, Joseph George (@jbgeorge), executive director for the HP Big Data Servers and Solutions group, shares details around their new Big Data Server family, the joint engineering effort with Hortonworks and what this means for Hadoop/Big Data community.

The Apache Hadoop project continues to surge ahead with significant contributions and key feature milestones.

But innovation comes in a number of forms.

And we at HP are challenging the status quo of Hadoop architectures – and purpose built big data servers just run Hadoop better.

Earlier this week, we announced the new HP Apollo 4000 Big Data server family:

  • The HP Apollo 4200 server, which is now the industry’s densest 2U server on the market, which can hold up to 28 LFF drives or 50 SFF drives – that’s 224 TB of direct attach storage in 2U.
  • The HP Apollo 4510 server, designed for object and cloud storage, which can hold up to 68 LFF drives in a 4U space – you can get 5.44 PB per rack. (That’s right – PETABYTES.)
  • The HP Apollo 4530 server, designed for hadoop and analytics, which has 3 server nodes in a 4U chassis, with 45 LFF drives, perfect for Hadoop triplication  

This family of compute solutions is enabling our customers to become data driven organizations – to know their business better, to serve their customers more effectively, and to manage more data with less space, less power, and more overall capacity.

On top of that, the HP Server group took it a step further with the innovative HP Big Data Reference Architecture – an architecture design that disaggregates compute and storage, using purpose-built products like HP Moonshot for compute and HP Apollo 4500 servers for storage.

And we were proud to partner with Hortonworks to develop the YARN Labels. Why is this important? These labels allows users to create pools of compute nodes where applications run, so it is possible to dynamically provision clusters without repartitioning data. Most interesting is that with labels, we can choose to deploy the Yarn containers onto compute nodes that are optimized and accelerated for each workload. Not only did we create, we’re also contributing it back to the community.

The result was a more performant Hadoop cluster that was able to get 2x the performance in 50% of the datacenter space.  Plus customers are now able to scale compute or storage independently, allowing for them to be intelligent on Hadoop cluster design, based on whether their data was accessed frequently (hot data) or not accessed very much at all (cold data). And due to the disaggregated nature of this architecture, the Yarn Labels feature was able to dynamically allocate only storage nodes or only compute nodes, and we were able to be specific about which nodes that they were allocated to.

Together, we have developed a more intelligent Hadoop scalable architecture – and it’s helping our customers achieve a 360-degree view of their business.

Wait – all this from a server solution vendor? You better believe it.

At HP, we are proud participants in the Apache Hadoop project, not sideline spectators.  Working with customers to understand their big data needs, building server solutions that address their big data challenges, making software contributions to the open source community, and pushing the boundaries of what we as a Hadoop community can do – it’s what we do.

And it’s apparent in the ingenuity and forward thinking built in to the new HP Apollo 4000 server family.

But this is just the beginning. Expect to hear more from HP and Hortonworks to provide you some real-world use cases on how you can apply our purpose built big data servers for your modern data architecture.– we’re looking at more ways to make your Hadoop experience more performant, more cost-effective, and more insightful.

In the meantime, check it out for yourself – learn more at

Let’s keep pushing the limits.


Leave a Reply

Your email address will not be published. Required fields are marked *