[bug] Txs in mempool are sometimes not mined into upcoming block

[ch4ot1c]

Using both sendtoaddress and create / sign / sendrawtransaction, transparent transactions are successfully created accepted into the mempool by the daemon. However, they often remain in the pool after blocks are mined, without being marked invalid or otherwise used or removed.

This happens exclusively on the testnet.

The BTCP team has spent some time investigating this 'sometimes' behavior, with focus on testing for regressions. None are apparent, so it appears there is a logic issue within.

We should run testblockvalidity with the contents of the mempool while it is full of candidate txs. CreateBlockTemplate also neesd to be investigated, to find the reason that some txs (txs are sometimes?) being accepted into upcoming blocks.

Don't forget to run the daemon with -debug for additional logging output.

---

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.