EPIC: Refactor existing e2e testing framework

[MSalopek]

Problem

ICS e2e testing uses a long bash script to initialize the testing environment.

Although there is nothing wrong with using bash scripts they quickly become poorly maintained due to their imperative and verbose nature.

• the script

Besides that, in the current setup it is difficult to achieve the following:

1. using arbitrary provider and consumer chains
2. configuring multi-consumer setups
3. getting comprehensive and understandable test output
4. import the e2e testing framework into both provider and consumer chain repositories
5. run in the CI

Closing criteria

At minimum the new test initialization should become declarative and allow easier running in the CI and allow arbitrary chain binaries to be used.

Related:

• Refactor and simplify e2e testing startup gaia#2434
• Refactor and simplify e2e test traces loading and execution gaia#2435
• Refactor existing e2e tests to take traces as inputs #875

Tasks

Beta Give feedback

1. make e2e it's own module (de-couple from ICS lib and versioning scheme) #1132
   scope: testing type: tech-debt +2
2. e2e tests fail with --verbose #1167
   type: bug +1
   
3. reduce number of steps in e2e happyPath #1178
   scope: testing type: tech-debt +2
   
4. update e2e initialization #1133
   S: Productivity dev-ops scope: testing type: refactoring type: tech-debt +5
   
5. Evaluate e2e testing tooling landscape #1134
   S: Productivity dev-ops scope: testing type: tech-debt +4
   
6. Automate the base case e2e infrastructure creation #1135
   scope: testing type: tech-debt +2
   
7. evaluate benefits of adding e2e tests of other libs into our CI [cosmos-sdk, ibc-go, other deps]
8. CI: refactor how e2e tests are run in github actions #1175
   good first issue scope: testing type: tech-debt +3
   
9. Support backward compatiblity checks in e2e testing #1559
   S: Productivity dev-ops scope: testing type: tech-debt +4
   

[mpoke]

@MSalopek Is this EPIC still needed?