Bidyut Bhattacharjee created HIVE-12873:
-------------------------------------------
Summary: Saving, sharing and securing Hive visualizations
Key: HIVE-12873
URL: https://issues.apache.org/jira/browse/HIVE-12873
Project: Hive
Issue Type: Improvement
Components: Authentication, Authorization, Hive, Logging, Statistics
Environment: HDP 2.3.2 on Hortonworks Sandbox, Windows OS, Hive
Visualization in general.
Reporter: Bidyut Bhattacharjee
Assignee: Thejas M Nair
We can create charts from Hive tables using the Visualization tab on Hive Query
Editor->Visualization, on the HortonWorks Hadoop Ambari Web Interface. The
'data explorer' tab and the 'advanced visualization' allows us to create
slices of the data
Currently, the visualizations/reports created under the 'Visualization' tab
cannot be saved.
I would like to request the following features to be added to Hive
Visualization. In my opinion the following features will boost product adoption
as reporting is a critical component of data-centric environment.
1. Ability to SAVE the HIVE visualizations
Users should be allowed to save the visualizations or reports created through
the 'data explorer' and 'Advanced visualization' features.
2. Ability to SHARE the saved HIVE visualizations or reports
After the visualizations or reports saved users should be able to share the
reports with other users through URL access.
3. Securing the saved visualizations/reports
It is important to secure the saved reports.
The following reporting 'roles' may be considered for controlling access to the
reports and to manage the reports:
(3.1) Report Administrator: The 'Report Administrator' administers(backs-up
reporting metadata, restores reports and metadata, will do performance tuning,
provide access rights, creates reporting roles, maintains and manages reporting
system wide configurations etc) of the overall reporting environment.
(3.2) Report Publisher: The 'Report Publisher' role is created by the 'Report
Administrator'. The publisher would be responsible for creating reports or
visualizations for the 'Report Browsers'. The report publishers would be able
to alter the report metadata of reports they have access to.
(3.3) Report Browser : The 'Report Administrator' and the 'Report Publisher'
should be allowed to create the 'Report Browser' role. The 'Report Browser'
browses or consumes the reports.
All these roles should be available by default (template) , custom roles may be
derived from the these roles.
4. Emailing reports
Report Publishers or Report Administrators should have the ability to schedule
automated notification the report consumers that the 'reports' have been
updated after scheduled data refresh executed successfully, this can be done by
simply emailing a hyperlink of the specific reports or by sending an SMS (Short
Messaging Service) to a mobile device. Alternatively, 'Report Browsers' should
be able to 'Subscribe' to the reports to be delivered to them at a scheduled
interval ( hourly, daily, weekly, monthly, quarterly or yearly)
5. Exporting reports
Once the reports are created or saved the users should be able to export the
reports to one of the following formats:
5.1 PDF (Portable Document Format)
5.2 Microsoft Excel ( .xls or .xlsx extension)
5.3 LibreOffice or OpenOffice (.ODS) format.
6. Dashboards
Dashboards contain multiple reports. Once the visualizations are saved, users
should be able to create dashboards using the saved reports. Example: Let's
assume that we are able to save two separate visualizations or reports. Report
'Monthly Sales' consists of 'sales by month' and the report 'Sales by state'
consists of 'Sales by each state'. Users should be able to create a sales
dashboard 'Sales Dashboard' comprising of both the reports. The ability to
display separate visualizations into one dashboard is very useful as it allows
re-usability, effective for decision making, saves time and improves analyst
productivity.
A 'Dashboard' tab may be added to the Hive->Query Window web interface.
7. Report execution statistics
There should be a repository to log every report execution. The log table/file
should store the unique execution identifier, date and time the report or the
dashboard was executed, duration for rending the report in seconds, the user
who executed the report, error message if any, report path, calling
application name etc.
8. Automated report documentation
Most reports in the corporate environment suffers from lack of documentation.
The report documentation should be created automatically. The following basic
information should be available in the documentation for each report or
visualization.
(8.1) Created by user identifier
(8.2) Created on
(8.3) Report description
(8.4) Report fields and data type for each field
(8.5) Description of or each field
(8.6) Report update date
(8.7) Report updated by
(8.8) Report update comments
Please feel free to drop an email if you have any questions.
Thanks for your help and consideration in advance.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
