RBF logic fails to reap stale outstanding TXs if RBF limit had been reached

[kantai]

A testnet miner ended up stalling seemingly due to an issue in the RBF logic.

I believe the relevant code is here:

https://github.com/blockstack/stacks-blockchain/blob/67dccdf6d1c28e482d3d71eb616cf8682dac863d/testnet/stacks-node/src/burnchains/bitcoin_regtest_controller.rs#L1033-L1073

In the first code block: we check if an outstanding TX has been committed.
In the second code block: we check if we're already exceeding the RBF limit.
In the third code block: we check if we should abandon any outstanding TXs because they've been reorged or stale.

I think that in the following scenario:

• Outstanding TXs cause the miner to reach the "RBF limit"
• Those outstanding TXs are reorged or never mined

The second code block would prevent the miner from ever getting to the third block and therefore some stale outstanding TXs would prevent the miner from ever producing a new commitment.

[kantai]

Note: If you are experiencing this issue, restarting the node (no need to wipe chainstate) would resolve the issue

[diwakergupta]

Fixed by #2694

[blockstack-devops]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.