About Release 7.6.1
This site contains documentation for HPE Ezmeral Data Fabric release 7.6.1, including installation, configuration, administration, and reference content, as well as content for the associated ecosystem components and drivers.
7.6.1 Installation
This section contains information about installing and upgrading HPE Ezmeral Data Fabric software. It also contains information about how to migrate data and applications from an Apache Hadoop cluster to a HPE Ezmeral Data Fabric cluster.
7.6.1 Data Fabric
HPE Ezmeral Data Fabric is the industry-leading data platform for AI and analytics that solves enterprise business needs.
- HPE Ezmeral Data Fabric File Store
  HPE Ezmeral Data Fabric File Store is a distributed file system for data storage, data management, and data protection. File Store supports mounting and cluster access via NFS and FUSE-based POSIX clients (basic, platinum, or PACC) and also supports access and management via HDFS APIs.
- HPE Ezmeral Data Fabric Object Store
  The HPE Ezmeral Data Fabric Object Store is a native object storage solution that efficiently stores objects and metadata for optimized access.
- HPE Ezmeral Data Fabric Database
  HPE Ezmeral Data Fabric Database is an enterprise-grade, high-performance, NoSQL database management system that you can use for real-time, operational analytics.
- HPE Ezmeral Data Fabric Streams
  HPE Ezmeral Data Fabric Streams brings integrated publish and subscribe messaging to the Data Fabric Converged Data Platform.
  - Architecture
    Streams contain topics that have logical collections of messages.
  - Producers
    Producers are data-generating applications, such as sensors in automobiles or activity loggers in servers. Producers create messages with the collected data and publish the messages to HPE Ezmeral Data Fabric Streams topics, specifically, to HPE Ezmeral Data Fabric Streams topic-partitions.
    - How Messages are Published
      To publish a message, a producer sends a record to the producer client library, which batches the records before sending them to the server.
    - Modes of Publishing
      Describes different modes of publishing.
    - How Partitions are Chosen for Messages
      Since the number of partitions in a topic can change over time, producers regularly refresh the information that they have about the topics that they know. This refresh interval is controlled by the metadata.max.age.ms configuration parameter.
  - Consumers
    Consumers are applications that you create such as analytics applications, reporting tools, or enterprise dashboards.
  - Stream Replication
    You can replicate streams to other Data Fabric clusters worldwide, or to other streams within a Data Fabric cluster.
  - Stream Security
    The adminperm, copyperm, comsumeperm, produceperm, and topicperm security permissions protect topics in a stream from unauthorized access. In addition, data-fabric supports user impersonation.
- HPE Ezmeral Unified Analytics
  Describes the HPE Ezmeral Unified Analytics Software and provides a link to more information.
- Kubernetes Interfaces for Data Fabric
  This section describes the Kubernetes Interfaces for Data Fabric, which include the Container Storage Interface (CSI) driver for multiple container-orchestration systems, and the FlexVolume driver for Kubernetes.
- Cluster Management
  Provides a synopsis of the various cluster components and their management.
- Performance
  Describes how to tune system performance, manage RDMA, and optimize CLDB tables.
- Security
  Provides an overview of the Data Fabric security features.
- YARN
- Client Connections
  The following sections describe how a client connects to local and remote data-fabric clusters.
7.6.1 Administration
This section describes how to manage the nodes and services that make up a cluster.
7.6.1 Development
This section contains information related to application development for Ezmeral ecosystem components and HPE Ezmeral Data Fabric products, including the file system, Database (Key-Value and JSON), and Event Streams.
Other Docs
This section contains release-independent information, including: Installer documentation, Ecosystem release notes, interoperability matrices, security vulnerabilities, and links to other data-fabric version documentation.
Glossary
Definitions for commonly used terms in MapR Converged Data Platform environments.

How Messages are Published

To publish a message, a producer sends a record to the producer client library, which batches the records before sending them to the server.

The producer client library sends the records to the server when any of the following conditions are met:

The producer client library has batched enough messages to make an efficient remote procedure call (RPC) to the server.
A message has been queued for the amount of time that is specified for the streams.buffer.max.time.ms configuration parameter.
For the Java client, the default interval for flushes is 3000 milliseconds. For clients based on librdkafka (for example, C, Python, and C#), the default interval for flushes is 0 (zero) milliseconds.
The producer client library has batched messages beyond the value of the buffer.memory configuration parameter.
The application explicitly flushes messages.

TIP The default number of threads used for flushing messages is 64. In most cases, this number provides excellent performance. However, you can adjust this number by setting a value for the fs.mapr.threads parameter in the core-site.xml file on your client node.