MapR Data Technologies - The 'Gloves Off' Series. 1 of 4 - Storage

Contributed by

6 min read

MapR Data Technologies is Unique and a ‘Must Evaluate’ for anybody seriously interested in driving Digital Transformation

In this series of four short articles, we set aside the more subtle art of ‘content marketing’ and shout from the rooftops about MapR’s World-Beating Storage, World-Leading Event Streaming, Best-in-Class NoSQL DB and the ultimate convergence of those three, the Global Data Fabric; a multi-environment, global namespace, in-place analytics data platform built to meet today's digital disruption head-on.

For anybody seriously interested in driving digital transformation and who has a genuine sense of urgency for enhancing their data infrastructure to best accelerate and sustain digital transformation, this short series of articles is for you. In this first of the series, we cover the bedrock for any data infrastructure; storage. Put your assumptions on storage aside because this storage is game-changing.

Top Reasons Why MapR Storage is Genuinely World-Beating

Global Namespace

  • Data in MapR is viewed singularly across any number or type of environments whether cloud, on-premise or at the edge. This is a genuine, single storage fabric and the only one of its type.

In-Place Analytics

  • The prevailing assumption is that for analytics, data has to be taken from NetApp or EMC or IBM or (insert your storage vendor here) and put somewhere else, today most likely Hadoop, to start a data pipeline for analytics. MapR is exabyte-scale, multi-temperature storage – with – in-place analytics; open access to all the compute and analytics innovation available; unified and interactive storage for both analytics and operations with the robust security and multitenancy to support this powerful new paradigm.

Just based on the above two unique capabilities, MapR customers enjoy reduced hardware costs, reduced deployment time and reduced operational complexity. Let’s continue…

Control Management

  • Object tiering both on- and off- premise with multi-temperature storage across SSDs and HDDs; the intelligent placement of data with global namespace and read-write access across all tiers. Providing unlimited control over storage cost-points from ‘hot’ to ‘cold’.

Deploy Anywhere

  • Take this MapR customer example; a global medical equipment company has MapR at ‘the edge’ (within MRI machines), as the data center (within the Hospital) and in the cloud for deep learning (cloud-based, multi-hospital clusters). If an organization needs to deploy across more than one cloud provider to spread risk or to provision the full range of tools or SLAs that it demands, for example, ALL of the data can be analyzed from a single logical view, in-place. A large telecommunications company runs MapR as small VMs at the edge and an automotive company bundles three Intel Nooks for edge data processing. Unlimited flexibility and agility, anywhere.

Extreme Scalability

  • MapR is arguably the most scalable data storage available today. Scale to 1000+ nodes and 100PB+ of data. The largest global companies in consumer electronics, technology, finance, retail and automotive rely on MapR for their business-critical, system-of-record storage and have chosen MapR over NetApp, EMC, IBM and their other historical storage providers. A direct comparison with HDFS is equally stark; MapR supports trillions of files. HDFS as part of the classic Hadoop ecosystem only supports 100M to 200M files which is insufficient when today’s storage must support a decade or more of expected data volume.

Always Available

  • MapR assumes failures are common as a baseline for availability. For one of MapR’s most global and strategic deployments, during SAP’s evaluation of most every major storage provider to underpin HANA Enterprise Cloud, they pulled every cord they could think of, every rack and generally tried to sabotage each storage candidate in every way possible. Only the most unlikely and extreme combinations led to any issues with MapR. Not so with the other candidates. It is unbelievably difficult to get MapR to fail and MapR was chosen. Note: 5 nodes are minimal; 3 nodes can fail and MapR will still be up and running. This is high resilience and reliability with global consistency.

Built-in Security

  • MapR has iron-clad security; industry-leading multi-tenancy, encryption, authorization and authentication as well as flexible access control expressions. MapR has the non-functional, business-critical-grade capabilities that enterprises expect from their most strategic systems, baked-in from the ground-up.

Interoperability Across Your Entire Enterprise Infrastructure

  • MapR exposes open interfaces to external clients. You can bring data into MapR using the simple NFS and/or POSIX access methods and expose the same data to your compute, analytics and processing tools of choice. NFS is a well-established industry standard open protocol which allows you to deploy both legacy as well as newer applications on the same storage platform. Using NFS to bring in data also eliminates the need to find and manage ETL tools just to ingest data, which further minimizes overall management.

MapR’s customers have been wowed by our storage capability. It completely re-writes both storage economics – and – how interoperable storage can, and needs to be, to underpin the operational analytics and intelligent application development that sit at the heart of today's corporate response to industry disruption and digital transformation.

In the second article, we will take a look at MapR’s World-Leading Events Streaming. While Big Data needs Big Storage, Big Data happens one event at a time. Event streaming capabilities underpin IoT deployments, in-the-moment analytics and the fluid data communications required for today’s microservices- and container-based applications.

This blog post was published January 23, 2018.