New Project Instance - V2

[markoprodanovic]

There have been a lot of small changes to code/documentation.

The version on W drive is quite outdated now.

We want to throw up a separate instance on Tableau Server to test the new version.

only Marko and Alison will have access

[markoprodanovic]

@alisonmyers could I get write access for the LA Data Projects folder?

[markoprodanovic]

@alisonmyers still getting some permission issues with Teamshare

Even moving module-progress to a subfolder within Module_Progress does not give Tableau Server the right access via UNC connection.
Should I reach out to Andrew again?

Another part of me thinks, if we have the blessing to do so, perhaps I can look into uploading to OneDrive via script and connecting from there? (#12) (#16)

..also looking into AWS more these days but still feels like I don't have a firm handle on it (but thinking how we can maybe utilize S3 buckets to host our CSVs in the future)

[markoprodanovic]

@markoprodanovic @alisonmyers

Resources:

(seems simple - OneDrive)
https://help.tableau.com/current/pro/desktop/en-us/examples_onedrive.htm

(seems more complex - AWS)
https://www.tableau.com/about/blog/2017/5/connect-your-s3-data-amazon-athena-connector-tableau-103-71105

[markoprodanovic]

Progress..

AWS is looking feasible! What I'm thinking needs to happen to make this work:

• single data output:
  • right now we output multiple CSVs and join in Tableau
  • let's eliminate any joining/merging in Tableau and do it all programatically
  • let's prompt user when running to specify whether they want user filters or not -> if so, a third join happens to include the course_entitlements.csv
• Host data in Amazon S3 Bucket (AWS)
• Set up SQL query to build a table based on this file with Amazon Athena (AWS)
• Connect Tableau using Amazon Athena - you will then be able to select database > table
• Add logic to end of script for upload to our S3 bucket (note that AWS can, i think, cache old versions of our dataset automatically (so no more zipping needed)

[alisonmyers]

@markoprodanovic
OneDrive - waiting for confirmation that this is a FIPPA compliant solution - no harm in doing a bit more research

Updating Teamshare - Module_Progress folder issues - let's replace and keep a version of old somewhere handy - I trust that the updates work.

• Stick a static version of the dashboard onto Tableau server with data pull on Tuesday
• replace Module_Progress with updated version including updated data
• just don't add permission to the "live" version until we confirm it works and remove the static version from the server

[alisonmyers]

Progress..

AWS is looking feasible! What I'm thinking needs to happen to make this work:

• single data output:

  • right now we output multiple CSVs and join in Tableau
  • let's eliminate any joining/merging in Tableau and do it all programatically
  • let's prompt user when running to specify whether they want user filters or not -> if so, a third join happens to include the course_entitlements.csv

• Host data in Amazon S3 Bucket (AWS)

• Set up SQL query to build a table based on this file with Amazon Athena (AWS)

• Connect Tableau using Amazon Athena - you will then be able to select database > table

• Add logic to end of script for upload to our S3 bucket (note that AWS can, i think, cache old versions of our dataset automatically (so no more zipping needed)

@markoprodanovic Test scenario - use the Canvas reporting bucket and see if we can get a live connection to Tableau server with that dataset.

[markoprodanovic]

Leaving this issue open as it touches on data storage conversations that are still happening