To ensure a baseline performance and to ensure basic functionality in the target deployment environment, manual testing needs to be conducted.
Tests to do:
| Time |
RAM (MB) |
| 1034 |
538.2MB |
| 1117 |
639.0MB |
| 1217 |
649.4MB |
| 1332 |
669.1MB |
| 1608 |
683.1MB |
| 1708 |
704.4MB |
| 1817 |
704.3MB |
| 1911 |
704.5MB |
| 2024 |
704.6MB |
| 0001 |
705.1MB |
| 0112 |
705.1MB |
While being a bit high for my liking, the phantombot also hovers about 700MB. This would be acceptable performance for now.
To ensure a baseline performance and to ensure basic functionality in the target deployment environment, manual testing needs to be conducted.
Tests to do:
GOAL: test if memory consumption is stable and is within acceptable levels (our target machine has 4GB of ram)
Results: I took several reading of memory consumption of the bot (BE) over several hours
While being a bit high for my liking, the phantombot also hovers about 700MB. This would be acceptable performance for now.
2. Long runtime, no load
GOAL: Test if the system is stable, and various Authkeys are correctly renewed
Results: issue found and temporarily resolved: Fix twitch oauth timeout #70
3. Long runtime, with high chat load
GOAL: Test system stability and performance while listening to a busy chat
4. Accessibility in prod environment
GOAL: Test if the panel + all APIs are resolvable while the system is in a prod-like configuration
RESULT: uris get resolved correctly when served on one port, with the panel served with nginx
5. Oauth authorizations working in prod environment
GOAL: Test if both twitch oauth client apps work when serving the app via one port, accessing it via a public domain
6. Clym chat "stress test"
Goal: migrate all commands to new bot, let chat try to break bot
7. Giveaway Beta
Goal: give everyone a bunch of coins, and run multiple giveaways to test the bot