Top 10 Reasons Customers Choose MapR

Not all big data platforms are created equal. Beyond the marketing claims, there are real differences which impact the bottom line while making IT operations easier. Here are 10 reasons why more customers are choosing MapR.

High Availability

1. High Availability

MapR takes a holistic approach to high availability. MapR architecture distributes NameNode metadata across all worker nodes in the cluster providing self-healing from multiple failures without requiring additional configuration or hardware. MapR allows instant recovery, with files and tables available rapidly after node failures or cluster restarts. Jobs on MapR do not have to be re-started on node failures and always run to completion. MapR also provides NFS HA for continuous uninterrupted access.

Learn more
World-Record Performance

2. World-Record Performance

Performance benefits businesses across multiple dimensions, not only in getting jobs/work done faster, but also squeezing more value from hardware. MapR has stellar performance achievements including the world record for TeraSort, MapR customer holding the world record for MinuteSort, MapR Database being 4-7x faster than HBase on other distributions, and OpenTSDB on MapR achieving 100 million data points/sec ingestion rates.

Learn more
Ease of Data Integration

3. Ease of Data Integration

MapR Direct Access NFS provides NAS like access to Hadoop. While other distributions provide poor-performing, read-only NFS, MapR provides complete random read-write capable, POSIX compliant, highly available, high performance NFS access for production use. With NFS, applications can stream writes directly into Hadoop, any language code works on Hadoop, standard linux commands and tools such as sed and awk are ready for use and your existing browsers and development tools work out of the box.

Learn more
Real Multi-tenancy Including YARN

4. Real Multi-tenancy Including YARN

Enterprise data hubs or data lakes require different users, applications to coexist on the same cluster with true job isolation and customized security. MapR is the only platform that is built to provide these capabilities with logical volumes, data placement control, and job placement control for both MapReduce v1 and YARN jobs.

Learn more
Complete Data Protection

5. Complete Data Protection

MapR ensures the same degree of backup and recovery capability offered by enterprise storage platforms. MapR Snapshots are guaranteed to be consistent, unlike in other distributions, so the snapshot accurately captures the exact state of the cluster at the time the snapshot was taken. Mirroring in MapR allows you to replicate data efficiently across clusters enabling data sharing between production sites, production and research environments or between on-premise and cloud infrastructures.

Learn more
Lowest Total Cost of Ownership

6. Lowest Total Cost of Ownership

With world-record performance under its belt, MapR is the most economical choice for building a Hadoop cluster. Supported by a No-NameNode architecture, MapR clusters are also homogeneous and therefore easy to maintain and scale unlike other Hadoop distributions which require special-purpose hardware or complex configurations. In addition, MapR provides fine-grained multi-tenancy to maximize system resources and support multiple workloads and distinct user groups efficiently.

Learn more
Enterprise-Grade NoSQL

7. Enterprise-Grade NoSQL

MapR Database, the in-Hadoop NoSQL database that uses the HBase API, was recently recognized as the top-ranked NoSQL key-value database for current offering. Its high throughput, consistent low latency, enterprise-grade features, and Hadoop integration allow you to deploy business-critical, real-time operational analytics applications. As MapR customer Atzmon Hen-tov of Pontis puts it, "MapR Database requires about half the machines compared to other [NoSQL] platforms. This dramatically reduces the cost of a new system."

Learn more
Unbiased Open Source

8. Unbiased Open Source

The MapR Distribution provides customers with more flexibility and choice for their open source projects. MapR supports multiple execution frameworks such as YARN and Spark, and multiple options for SQL-on-Hadoop technologies, machine learning packages, NoSQL databases. Furthermore, MapR uniquely supports backward compatibility across multiple versions of projects.

Learn more
Read-Write File System

9. Read-Write File System

Unlike HDFS, which follows the write-once-read-many paradigm, the MapR Data Platform delivers a true random read-write capable, POSIX compliant file-system providing unique features such as read-write NFS. These capabilities enable Hadoop as a real-time, enterprise storage and processing platform.

Learn more
Enterprise-grade Security

10. Enterprise-grade Security

MapR security includes wire-level encryption for Hadoop, granular authorization with ACLs and boolean access control expressions, and authentication via existing Kerberos infrastructure or a simplified username-password based mechanism. MapR also implements security projects from Apache for additional layers of protection.

Learn more