After a heavy load test with thousands of transactions from a fixed set of accounts (maybe relevant due to this question), the ready pool of some nodes never empties, and for this reason many transactions get lost.

It's a Frontier based PoA chain currently with 1 validator for Aura/Grandpa and the others are full nodes.

  • Block slot: 1s
  • BlockWeights: ~500ms
  • BlockGasLimit: 10M enter image description here

The test sends thousands of ERC-20 transfers via WebSocket and listens to their events in one of the nodes

In the logs of one of the nodes I can see that the ready queue ahs the 2812 txs:

2023-09-27 17:53:35.070 DEBUG tokio-runtime-worker txpool: Resubmitted. Took 132 ms. Status: PoolStatus { ready: 2812, ready_bytes: 940124, future: 818, future_bytes: 295298 }

The only way to clear the queue seems to be restarting the nodes.

edit: more info about the chain


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.