Best practices for Hadoop in production – Panel discussion facilitated by Forrester analyst
Listen to some of our customers discuss their success with MapR in production.

Watch

Testing Hadoop in development environments is dramatically different from deploying Hadoop into the data center with the rest of your enterprise architecture. Maintaining SLAs, reliability, security, and disaster recovery are critical. The MapR Converged Data Platform ensures production success with an architecture designed specifically for business-critical applications, seamless data access and integration, and the ability to run both online and analytical processing and applications reliably on one platform.



KEY FEATURES
MapR CTO%2C M.C. Srivas presents the MapR architecture Part 1 of 2

Trusted Architecture for Business-Critical Applications

Only MapR provides Hadoop high availability (HA) everywhere across ALL Apache Hadoop projects. The result? Applications and insights your business can rely on. MapR provides 99.999% high availability, consistent point-in-time recovery of data as well as disaster recovery. MapR has built redundancy at every layer of Hadoop including protection from node failures, MapReduce failures as well as data access point failures.

High Availability on MapR: Download the Technical Brief
MapR and the Easiest Access to Hadoop Data

Seamless Data Access and Integration

MapR is the only distribution that allows Hadoop to be accessed as easily as network attached storage (NAS) via NFS. This allows faster data management and system administration without requiring multiple steps to access and move data. Users can mount the MapR cluster and perform traditional file system operations such as ingesting data directly, randomly read as well as benefit from thousands of existing enterprise tools to directly access and manipulate data on Hadoop. There are no performance bottlenecks or append-only files that prevent users from seamlessly integrating Hadoop with the rest of the enterprise data architecture.

Lower Operational Costs

MapR provides the best production-ready distribution for Hadoop that reliably runs both online and analytical processing and applications on one platform. This allows you to run more applications on one Hadoop cluster, minimizing operational costs. The MapR architecture is optimized to handle files and tables as first-class citizens for applications with ultra reliability, consistent high performance, and manageability. MapR runs the largest single production clusters of Hadoop in the world via:
  • Linear scalability that goes beyond the 100 million files limit in HDFS
  • Distributed metadata architecture - scales to trillions of files and tables, allowing users to store and process thousands of petabytes on one cluster
  • Processing of files and tables in one distributed storage layer allowing for NoSQL and Hadoop applications to work seamlessly on one platform