I just started my 'occasional' full-node and I am watching the size of the mempool grow after full block synchronization - watch -n 3 'bitcoin-cli getrawmempool true |jq "length"' - it has around 4000 pending transactions. My 'continuous' full-node, however, has around 18000 pending transactions which seem to never become available to my 'ocasional' node.

Looks like that when a node starts up it will receive the new transactions that arrive at the mempool, but the old transactions that had been laying there are never 'synchronized'. How can I synchronize the mempool of a recently launched node with the one of its peers?

I could shut down my other full-node and copy past mempool.dat into my other node but this solution is very ugly as I need to shut down a long-running node with many connections.

  • Why do you think it matters?
    – Claris
    Feb 10, 2020 at 19:05
  • off the top of my head: fee recommendations with custom code, setting up your electrum server and proving up-to-date intormation, fee bumping transactions made on your mobile wallet through bitcoin core,...
    – Pedro
    Feb 10, 2020 at 19:26
  • Fee estimation wouldn’t work if you stuffed the men pool suddenly with old transaction. Electrum servers won’t be started and stopped frequently.
    – Claris
    Feb 10, 2020 at 19:30
  • Fee estimations made by looking at the current size of the mempool would work with old transactions (not bitcoin core recommendations, that's why I wrote 'custom code'). For example you could recommend a fee that is at the top 0.8 vMB of the mempool.
    – Pedro
    Feb 10, 2020 at 19:33
  • If you're writing custom code you can surely just not have an "occasional" node.
    – Claris
    Feb 10, 2020 at 22:06


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy