Degenerate System State Exploration
tylerwilding-wk opened this issue · comments
tylerwilding-wk commented
The idea here is that we should have at least some explicit exploration and categorization of system behavior under duress from edge case scenarios. Examples:
- What is the behavior of the transactor / transaction log, indexer / indexes with datoms that fall far outside of the contract we've designed for? (Eg, What happens when we put a few thousand 1MB datoms into the indexes? Can we come up with cases that can consistently crash our index processes?)
- What is the behavior of the system for very thrashy datoms?