Follows %TzTraffic deployed on Tznet

Part of a larger objective to "automatically test each new release and protocol in a realistic simulated environment"

Context: Previous milestones lead to:

• Building a realistic simulated environment : TzTraffic + TzNet
  • TzTraffic: PoC done, improvements coming but now needs specifics from tests to decide new features to implement
  • TzNet : - manual docker-compose based deployment on single node - Kubernetes deployment of nodes + client + baker ok (including scaling), however misses split of baker wallet.
    • now adding prometheus + grafana + teztale etc.

This Milestone focuses on three parts:

1. Deployment of the "realistic" simulated environment using Kubernetes and GCP

• TzNet deploys networks from 1 to 400 pods easily tztraffic#20
• TzNet pods include 1 "yes-node" + 1 "yes-baker" with a "yes wallet" to split stake ~equally tztraffic#21
• TzNet pods include prometheus + grafana + teztale archiver tztraffic#22 (closed)
• 1 TzTraffic is deployed on the network and injects operations correctly tztraffic#23
• 1 Teztale server is deployed and its bdd is backed up after the experiment tztraffic#24
• Logs, grafana and teztale data is backed up tztraffic#25
• TzNet is deployable to GCP tztraffic#26

2. Describe, implement and manually run tests for releases

• Define release tests metrics and how to display them tztraffic#27
• Define an as-small-as-possible set of tests that provide defined metrics tztraffic#28
• Write a general document: What we want to check, how, at what infrastructure cost, with what constraints on release (in particular schedule)? tztraffic#29
• Write a document for each test, ideally make it executable/automatable ;) tztraffic#30
• Test previous releases tztraffic#31
• Test next release tztraffic#32
• Predict/Assess cost of running each experiment depending of its configuration tztraffic#33
• Document and publish (at least internally) a tutorial so that Release Management could manually run tests prior to any release with no help from people from the team tztraffic#34

3. Aggregate and analyze results from experiments/tests

• Develop grafana dashboard needed for aggregated data tztraffic#35
• Interact with teztale team to add dashboard for aggregated data if needed tztraffic#36
• Collect and save data for previous releases for comparison tztraffic#37

Predicted interactions:

• Release Management
• Teztale team
• Devops
• (Maybe DAL and Decrease Block Time team that work on related experiments)

Predicted follow-up goals:

• Automate deployment and tests
• Improve test scenarios, adding failures, heterogeneous behaviors, heterogeneous systems, etc.
• Automatically test each new protocol in a realistic simulated environment