[
https://issues.apache.org/jira/browse/RANGER-3923?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Madhan Neethiraj updated RANGER-3923:
-------------------------------------
Description:
Given the primary business value of Apache Ranger is to enable sharing of
resources, it will help if Apache Ranger provides an abstraction that enables a
set of resources/data across services, a dataset, to be the unit of sharing
instead of one or more resources in each service. This has several benefits,
like:
* A single policy to manage access to data in multiple services - like HBase,
Hive, Snowflake, Kafka, Google BigQuery, AWS S3, AWS Redshift, ADLS-Gen2. This
enables authorization to be centered around a purpose, like:
*
** Marketing Campaign 2022 dataset
** Sales 2021 dataset
** CA Claims 2021 dataset
* Enables different set of users to manage sharing data into a dataset and
manage access to the data in a dataset:
*
** Data owners share data into a dataset, with necessary masking, row-filters
and schedules; they can update the share details, including stop sharing into a
dataset.
** Dataset admins manage who has access to the data in the dataset. This
relieves data owners from having to micromanage access to the shared data, for
example when a user needs access to the data in multiple services to
participate in a project.
Attached document has more details on this new abstraction, including a number
of questions & answers that to help understand various aspects of this feature.
Please read and add your comments/suggestions.
was:
Given the primary business value of Apache Ranger is to enable sharing of
resources, it will help if Apache Ranger provides an abstraction that enables a
set of resources/data across services, a dataset, to be the unit of sharing
instead of one or more resources in each service. This has several benefits,
like:
# A single policy to manage access to data in multiple services - like HBase,
Hive, Snowflake, Kafka, Google BigQuery, AWS S3, AWS Redshift, ADLS-Gen2. This
enables authorization to be centered around a purpose, like:
* Marketing Campaign 2022 dataset
* Sales 2021 dataset
* CA Claims 2021 dataset
# Enables different set of users to manage sharing data into a dataset and
manage access to the data in a dataset:
* Data owners share data into a dataset, with necessary masking, row-filters
and schedules; they can update the share details, including stop sharing into a
dataset.
* Dataset admins manage who has access to the data in the dataset. This
relieves data owners from having to micromanage access to the shared data, for
example when a user needs access to the data in multiple services to
participate in a project.
Attached document has more details on this new abstraction, including a number
of questions & answers that to help understand various aspects of this feature.
Please read and add your comments/suggestions.
> Dataset policies
> ----------------
>
> Key: RANGER-3923
> URL: https://issues.apache.org/jira/browse/RANGER-3923
> Project: Ranger
> Issue Type: New Feature
> Components: Ranger
> Reporter: Madhan Neethiraj
> Assignee: Madhan Neethiraj
> Priority: Major
> Attachments: Apache Ranger - Dataset.pdf
>
>
> Given the primary business value of Apache Ranger is to enable sharing of
> resources, it will help if Apache Ranger provides an abstraction that enables
> a set of resources/data across services, a dataset, to be the unit of sharing
> instead of one or more resources in each service. This has several benefits,
> like:
> * A single policy to manage access to data in multiple services - like
> HBase, Hive, Snowflake, Kafka, Google BigQuery, AWS S3, AWS Redshift,
> ADLS-Gen2. This enables authorization to be centered around a purpose, like:
> *
> ** Marketing Campaign 2022 dataset
> ** Sales 2021 dataset
> ** CA Claims 2021 dataset
> * Enables different set of users to manage sharing data into a dataset and
> manage access to the data in a dataset:
> *
> ** Data owners share data into a dataset, with necessary masking,
> row-filters and schedules; they can update the share details, including stop
> sharing into a dataset.
> ** Dataset admins manage who has access to the data in the dataset. This
> relieves data owners from having to micromanage access to the shared data,
> for example when a user needs access to the data in multiple services to
> participate in a project.
> Attached document has more details on this new abstraction, including a
> number of questions & answers that to help understand various aspects of this
> feature. Please read and add your comments/suggestions.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
