Spike : Peer review of legal DB data reset tool - create tickets for refactoring/replacement

[JohnamLane]

Description:

Background:

The data reset tool created by Conrad is now deployed in Dev and Test. It is the sole means of programmatically setting up entities test data. It's been working great, but it has required some maintenance to match schema changes.

I propose this ticket be used to review the design of the tool to see if it can be decoupled from schema changes, or if it could be replaced wholesale with an alternative tool that is easier to maintain.

This ticket would be timeboxed and the winning proposal would entail it's own series of tickets.

Code is here: https://github.com/bcgov/lear/tree/master/data-reset-tool

Acceptance Criteria:

• Research the current state of the data reset tool. Get some ideas on what needs to be done.
• Facilitate Meeting to Determine what refactoring steps are needed to make the data reset tool resilient to data changes.
• Consider any process changes required to update the tool for future schema changes.
• New design would ideally handle multiple concurrent calls (current tool crashes at 3-4 requests at the same time)
• Add tickets to bring the work in to future sprints.

Design suggestion:

• Use db -> CSV mechanism for data manipulation. (Postgres copy)
• Look into notebooks for doing this.

[JohnamLane]

@shahriar-khan-fw @rstens This has gone up in priority since we last spoke because the tool currently doesn't support 'Add Detail' or colin event IDs

[sienna-oldaccountdontuse]

Peter: "Riyaz and I are looking into that, I am looking into different ways to map a CSV but it seems that we'll likely be using Thor's suggestion... but regardless I need to see the overall testing strategy from Riyaz"

[JohnamLane]

@Kaineatthelab @peter-freshworks Peter, Can you please clarify how the testing strategy information is currently related to your work of defining next steps in redesigning the data reset tool? As described above, this tool is the sole means of setting up test data for the legal API. It is used today by both the newman suite and the nightwatch suite. The current functions and interfaces of the API should remain the same. Are there any other details that you're missing?

[peter-freshworks]

@JohnamLane in our meeting about blockers Riyaz had some input about this, and so we added him on the ticket. I may know how the tool currently works and how to run it locally etc. but I do not have very much experience/knowledge with how QA uses it, the pipelines etc. and so I thought that might be important to know.