Please enable JavaScript.
Coggle requires JavaScript to display documents.
Connections MVP, For permanent removals, we need to clear up the whole…
Connections MVP
-
Focus 2 - Discrepancies
-
As with Current connections, we will rationalise columns, remove Programme membership history and interaction removed.
The bulk of our work will be to ensure the data is correct, based on the previous conversations around logic etc…
-
In Connections, we will hide the tabs Historic Connections, All and Hidden.
We don't want users to use Connections till the data is clean
- Remove interactivity first
- Focus on getting the data right
- Then reintroduce interactivity
-
For permanent removals, we need to clear up the whole thing, not just the Front End UI
-
-
Ran out of time to complete. Question we finished on was:How do we minimise the inability of Reval users to use the Connections function (currently on legacy Reval)?
1. Remove interactivity on new Reval
2. Remove access to legacy Reval
3. Data migration
4. Reinstate interactivity
- Tidy up presentation of data / columns etc
Can 1-4 above be done relatively quickly - days rather than iterations?
Could we extract FE improvements
Extract BE logical stuff - to support FE (do this collaboratively)
What do we need to hide / see?
We have things that are broken - we need to redevelop messaging and integration flows
|
|
\/
Then / Or should we create the Epic, and User Stories and work off those.
We have a template of what we want to see, but are we clear on what we need to migrate?
Note data on New Reval is very stale - will bad data negatively affect the work we need to do.
GMC doing an update to their Test environment for a week, commencing 10 Oct.
We should have access to clean data after that.
Tech improvement / Infrastructure considerations:
1- API gateway that needs updating? Y/N? Upgrade to the latest module version
This makes adding any endpoint or user requests possible - keeps the website available to users
Do at the first point we change any endpoint
2- Rebuild pipelines to Production (part of the Git repo set up). Something kind of works, but not well enough. At the same time, configure Github environments
See some of the work AndyD has been sketching out on reusability of workflows.
Address as soon - just before - we want to deploy to the Prod environment
3- DocumentDB database to replace single node db
Makes the DB MUCH less likely to break. Speed at which data can be updated in Reval is currently limited. When it crashes it impacts TIS as well.
Address before data migration
4- Monitoring and wider observability considerations
Fold this work into all the work - monitoring best addressed as you're doing work, rather than afterwards. e.g. as we take updates from GMC, start monitoring failed transactions
5- Redevelop messaging and integration flows
Data in Connections is correct / up to date
Do this first so we have something solid to test against? Slice it alongside the relevant bit of logic? Focus on queue that listens for updates
