Closed NicolasMahe closed 4 years ago
Seems valid, could you run the perf tests and check the result on the grafana to see the impact of such a modification, the before/after visualization for let's say 1000 executions would be great
Seems valid, could you run the perf tests and check the result on the grafana to see the impact of such a modification, the before/after visualization for let's say 1000 executions would be great
Same speed on the e2e tests with 10 executions: Before:
--- PASS: TestAPI/execution/many_executions_in_parallel (1.65s)
--- PASS: TestAPI/execution/many_executions_in_parallel/create_executions (0.60s)
--- PASS: TestAPI/execution/many_executions_in_parallel/check_in_progress (0.02s)
--- PASS: TestAPI/execution/many_executions_in_parallel/check_completed (1.03s)
After:
--- PASS: TestAPI/execution/many_executions_in_parallel (1.59s)
--- PASS: TestAPI/execution/many_executions_in_parallel/create_executions (0.56s)
--- PASS: TestAPI/execution/many_executions_in_parallel/check_in_progress (0.02s)
--- PASS: TestAPI/execution/many_executions_in_parallel/check_completed (1.02s)
Fixes https://github.com/mesg-foundation/engine/issues/1693
This PR make the lcd and rpc clients to only increase the account sequence when the transaction is ok.