Turns out that mainnet invariant testing represents the largest amount of usage. We talked about switching to a sub model instead of querying every block. In short, we’d subscribe to these events:
add/remove liquidity
remove withdrawal shares
open/close long
create checkpoint
In addition, we would randomly check the state 1/hr or so to run the invariant tests
Turns out that mainnet invariant testing represents the largest amount of usage. We talked about switching to a sub model instead of querying every block. In short, we’d subscribe to these events:
In addition, we would randomly check the state 1/hr or so to run the invariant tests