MapR Technologies, Inc.
MapR is a complete Distribution for Apache Hadoop that includes HBase, Pig, Hive, Mahout, Cascading, Sqoop, Flume and more. MapR’s distribution is 100% API compatible with Hadoop (MapReduce, HDFS and HBase). The M5 Edition includes advanced high availability and data protection features such as JobTracker HA, Distributed NameNode HA, Snapshots and Mirroring. The M5 edition, available on a subscription basis, is fully supported for the most demanding enterprise mission critical deployments.
High Availability. With JobTracker HA and Distributed NameNode HA™, MapR’s Lockless Storage Services™ architecture was designed with no single points of failure. Distributed high-availability is also built-in throughout the entire system. The M5 Edition’s JobTracker HA prevents lost jobs and painful restarts. Our Distributed NameNode HA™ eliminates one of the most frustrating failure modes of all other Hadoop distributions and provides both scale and performance advantages.
Mirroring for Business Continuity.
The M5 Edition makes disaster recovery easy and built-in. Going far beyond replication, MapR’s mirroring means you can set policies around your Recovery Time objectives (RTO) and mirror your data automatically within your cluster, between two clusters or between two data centers.
Protect your data and set point-in-time recovery with snapshots. Replication cannot protect data from application errors or deletion. Serious applications require the ability to recover to a specific point in time. MapR’s Snapshots require very low overhead and can be done on a frequent interval. Recovery is as easy as copying a directory or files from the snapshot directory to the current directory. You can also drag and drop from any file browser supporting NFS.
Share the cluster safely with multiple jobs. MapR M5 Edition protects the core system by isolating it from user jobs so runaway jobs can’t bring down your
entire cluster. Safely sharing the cluster also means higher utilization. Setting quotas for users and then tracking capacity usage is made easy whether you are managing one node or thousands. Automated alerts provide a built-in assistant for you to rely on as you manage the Hadoop environment.
Provision Jobs, Users and Groups and perform capacity planning. MapR is the first distribution to implement Volumes coupled with the ability to provision around users and groups. Volumes allow you to frame your application’s business requirements in Hadoop. Elements such as application Recovery Time Objectives, Recovery Point Objectives, required CPU and data resources, expected capacity growth and security requirements can all be defined at the volume level.
Revolutionary Direct Access NFS™ for realtime dataflows. MapR enables you to simply mount the cluster as an NFS volume giving Hadoop a full read/write storage system to support multiple readers and writers as well as full random read and write. Our patent pending NFS access allows applications to load data directlyinto the cluster and gain realtime access to the results.
Instant intuitive insight from MapR Heatmap™. MapR provides insight, complete access and power tools to help you understand and operate your cluster
with ease. Intuitive views into the most critical issues are automatically provided along with tools to easily provision a single node or thousands. Our graphical and programmatic interfaces are designed to scale to the largest clusters, and of course, we provide CLI and REST access as well.
MapR M5 Edition Support
- 24x7 phone and email support
- On-demand patches as needed
- Online incident submission and response
MapR M7 provides ease of use, dependability and performance advantages for NoSQL and Apache Hadoop™ applications. MapR M7 has removed the trade-offs organizations face when looking to deploy a NoSQL solution. M7 provides scale, strong consistency, reliability and continuous low latency.
MapR’s patented technology increases the speed and scale of Hadoop via a distributed read/write layer that eliminates file processing limitations, Java issues and the overhead and restrictions associated with a single name node. MapR’s underlying storage architecture reduces network overhead with automatic and transparent compression and enables highly optimized shuffle operations for further performance advantages. Because of these technical advancements, MapR excels at random I/O throughput with faster performance than any other Hadoop distribution. Greater hardware utilization provides many benefits, the most apparent being a strong ROI that encompasses hard dollar CAPeX and OPeX savings. This performance benefit is not at the sacrifice of uptime or ease of use but in addition to them. MapR fully utilizes your network and leverages Cisco’s UCS hardware to its full potential, delivering outstanding value.
MapR M7 is architected from the ground up to deliver reliability and performance without requiring compactions or background consistency checks to work smoothly. M7 provides dramatic scalability advantages with support for up to one trillion tables across thousands of nodes. M7 also provides instant recovery from failures, ensuring 99.999% availability for Apache HBase™ and Hadoop applications.
With M7, there are no RegionServers, additional processes, or any redundant layer between the application and the data residing in the cluster. M7’s zero-administration approach includes automatic region splits and self-tuning with no downtime required for any operation including schema changes.