What's New in Version 5.2
Provides a summary of new functionality in the 5.2 release
The 5.2 MapR release supplies new features for MapR-DB and MapR-Streams as well as debuting the MapR Monitoring tool that supplies deep monitoring capability across all of your clusters. For Ecosystem components, this release provides breakthrough improvements in deployment, supportability, and stability of Ecosystem components through the new MapR Expansion Pack (MEP) process.
MapR Monitoring
MapR Monitoring (part of the Spyglass initiative) provides the ability to collect, store, and view metrics and logs for nodes, services, and jobs/applications.
Administrators can use dashboards to visualize the metrics and logs. Dashboards can be customized to include metrics across clusters and nodes. Dashboards can also be customized to display metrics for specific services (YARN, Drill etc.). The MapR Community Exchange provides sample dashboards that administrators can use to get started.
MapR Monitoring Architecture describes the open-source components that MapR Monitoring uses to collect, aggregate, store, and visualize metrics and logs. Note that Grafana and Kibana are community supported. For more information about MapR Monitoring, see Using MapR Monitoring (Spyglass Initiative).
MapR-DB
- JSON Table Replication
This feature supports table replication for JSON tables. To specify the table type, use either the MapR MCS via the Create Table window option, the
maprcli table create
command with the-tabletype
parameter, or the MapR-DB JSON Java API Library.This feature also supports column-level replication for JSON tables by expanding the
-columns
option for the following commands:maprcli table replica add
maprcli table replica autosetup
maprcli table replica edit
- Enhancement for Querying JSON Tables
- If the field path that you specify for returning partial documents by using projection
does not exist in a document, the query returns the value of the
_id
field for that document. Also, projecting on the_id
field returns only the values of the_id
field in the documents. - JSON Table GET/SCAN Behavior Changes
- Reading with projected columns will not filter rows which do not have the projected
columns. For columns that the user does not have read access to, those columns are
assumed to be non-existent in the documents. Since access control over the row-key is
not available, even if the user does not have read permissions on any column in a table,
the row-key can still be read. This behavior change is applicable to the
find()
andfindById
methods of aTable
object. - Indexing MapR-DB Binary Tables with Elasticsearch
- MapR-DB 5.2 is compatible with Elasticsearch version 2.2, which you can get from https://www.elastic.co/.
MapR Streams Features and Enhancements
- Administrator APIs
- The Java Admin API has the following administrator functionality:
- Create topic and stream properties
- Delete topics and stream properties
- Edit topic and stream properties
- Manage topic and stream properties
- Spark-streaming Integration with MapR Streams
- Spark-streaming is now fully supported with MapR Streams for production use. For more information, see Integrate Spark with MapR Streams.
- Operating System Support
- Windows 64 bit and Mac client support has been added.
MapR-FS Features and Enhancements
- Hard Links on Files
- MapR 5.2 includes support for creating and retrieving hard links, which are directory entries that associate new names with existing files on the file system. For more information, see Hard Links.
- Extended Attributes on Files and Directories
- MapR now includes support for adding and retrieving extended attributes, which allow user applications to associate additional metadata with a regular file or directory. For more information, see Extended Attributes.
- Impersonation for all Users
- MapR now includes a new type of ticket, "servicewithimpersonation," that extends impersonation capability to all users. Any user with this ticket can now impersonate other users except the mapr superuser. For more information, see Generating a Service with Impersonation Ticket.
- Enhancements
- MapR 5.2 also includes:
- New parameters for FUSE-based POSIX clients to configure timeout values for attribute cache and entity cache. For more information, see Configuring FUSE-based POSIX Client.
- Support for running
df -h
to determine the amount of used and available disk space. For more information, see Determining the Amount of Free Disk. - A new property for reducing the failure detection time for Hadoop and FUSE-based POSIX clients. For more information, see Reducing the Failure Detection Time for File Clients.
- Automatic detection and distribution of all copies of a container if all copies end up on the same rack. For more information, see Determining Replication Factor.
MapR Expansion Pack (MEP)
MapR 5.2 supports MapR Expansion Packs (MEPs). A MEP provides a set of ecosystem components that work together. These components are fully supported and tested for interoperability. For more information, see MEP Components and OS Support.