Instructor Meeting 02.04.24 #2
Description
- Project is already implemented: that's fine
- We get 2 clusters
We can start with:
- Containerizing
- We should use Ingress because we will meet it later in class 9/10 (instead of API Gateway)
- Helm (is going to be handled in the class)
TODO:
- Role assignment: How do we want to split the work (which chief responsibilities)?
-- Spring Boot
-- Testing
-- Maven - Development: Should scale with storage
-> Adding Content Addressable Storage to the storage - Add something to handle the photo upload (on mobile devices)
- Think about on what to focus on (Start-Up Marketing-wise: f.e. Gamification)
-> reason about on which part of the product we want to focus on - Translate GitLab Pipeline to GitHub Pipe
- Replace Basic Auth, for example with the GitHub Plugin
- Don't deploy the Ingress yet, SSO is handled in the class later
Think about: How do we measure quality?:
- application gathers user statistics
- How can we know if everything runs correctly?
-> Helm health report should be added (will also be handled in class)
Second Diagram:
- Up to us: we can decide if we use the GitHub Pipelines or not
- It's good to try both possibilities and evaluate for us which one suits us better
DB (but applies to everything):
- Prioritize (HA-Failover for DB will probably not be needed for the course/to deliver MVP)
When would we need the Kubernetesses?
- Start local, go for docker-compose
- Then check for the auth feature
- Check if logging is implemented
- Then we can wait until 3rd Week of April to get Kubernetes
Product pitch in the end:
- Write why this idea is going to work? (convince the audience)
- For the final presentation: Make a run through with screen recordings (not a live demo)
- Story Line has to be consistent
How would we monetize our application?
- Sponsorings
- Sell Data to companies/restaurants
- Have a B2B Setup to let people use sponsered products
Put a Slack Team Channel in Slack ✔️
Ideas but not mandatory
- Optimize development to delivery time
- Blue/Green DBs ? - Schema Migration - Rollouts
- ArgoCD vs Flux play around
- Handle state
- Go crazy - service mesh (probably not)