Tableau Admin Scripts

This is a collection of scripts for features missing from the Tableau Admin suite of tools. It's focused on

archiving unused reports
emailing extract owners of failed extracts
parsing Tableau Desktop logs for queries
exporting workbooks and datasources in XML without attached data
Subscription framework that allows for more flexible report subscriptions

For all of this, a huge thanks to Ryan Mason for going through my crappy code, improving it everywhere, and writing 95% of the documentation

Initial setup

Access our helper functions.

export PYTHONPATH=CHECKOUT_DIR/libraries/

export PYTHONPATH=$(pwd)/libraries/

Dependencies

virtual environments

Setup virtualenv to handle libraries.

http://docs.python-guide.org/en/latest/dev/virtualenvs/

Postgres

Tableau Server is backed by a Postgres server. In order to be able to get information about the state of your server, the library requires that you have Postgres drivers installed. You'll also need to have a user that can access the system tables. Read this thread for an idea of how to go about this: http://community.tableau.com/message/340714#340714

Python Postgres Driver Downloads:

- Windows: http://www.stickpeople.com/projects/python/win-psycopg/
- OS X: [Homebrew](http://brew.sh/), `brew install postgres`
- Ubuntu: sudo apt-get install python-psycopg2

tabcmd

If you'd like to take actions on the server, you'll need a local copy of tabcmd to interact with the Tableau Server. To get tabcmd for mac/linux it's now pretty trivial on 9.0. Take a look at this post: http://community.tableau.com/message/343118#343118

Catching emails in development

If you'd like to see the emails that these scripts send, try using Mailcatcher or a similar tool. Mailcatcher will act as an SMTP acceptor and will show you the messages in a browser. This is a great way to see what you're going to send before sending to everyone.

Settings

All settings for these scripts live in libraries/settings.py. This file has constants used throughout these scripts. Not every setting is used by every script.

We recommend against checking this file in. We have included libraries/settings.default.py for reference. You can use it as a starting point for your settings with $ cp libraries/settings.default.py libraries/settings.py.

Running

Each script can be run separately. They share a common set of libraries and settings in the libraries/ folder.

archive_unused_reports

Find all the reports that are longer than settings.ARCHIVE_WINDOW. Alert owners, delete them.

    $ python archive_unused_reports.py

extract_error_email

Find all extracts from the last settings.EXTRACT_ERROR_WINDOW_HOURS hours. Email the owners the error messages.

    $ python extract_error_email.py

client_log_parsing

Find queries in Tableau Desktop logs.

tableau_export_workbooks

Export only the XML from Tableau Workbooks and data sources, and tries to insert the datasources into a database. Currently set up for Vertica. Needs to be abstracted for more export options

    $ python tableau_export_datascources.py

subscriptions

This is a full framework for emailing reports to users. Some of the features:

- Email multiple users
- Each email may contain multiple dashboards from multiple different workbooks
- (Slightly) intelligent caching. If a report is going to many users, it only pulls it down once.
- Filtered subscriptions. This solves for: User A is only supposed to see data for California, and User B data for Wisconsin
- Dynamic Filters (Sort of). Ex: Report can always be filtered to show only last saturday's data.

This is complex enough that it has its own readme in the subscriptions directory (readme.txt)

##queries

Collection of queries against the tableau postgres database that have been found to be useful.

Tableau Server REST API

You can interact with the Tableau Server via the Rest API

http://onlinehelp.tableau.com/v8.2/server/en-us/help.htm#rest_api_ref.htm

Python example REST

http://onlinehelp.tableau.com/samples/en-us/rest_api/example.py

ranjith520/tableau_admin