bitcoind should write to debug.log when it receives a shutdown signal so operator knows signal was sent while debugging/setup and for general logging.
Story
As a new bitcoind operator, it was not evident to me why bitcoind was running so slow. This manifested in several ways:
- RPC calls from localhost took 1-2 mins
- uptime went through the roof
- shutdown also took 1-2 mins
I have since realized the obvious (now that I understand bitcoin better) that I'm still verifying my own block chain. (This is taking freaking forever! I know I could use a lightweight...)
Rational
What I've proposed is a simple change that would significantly help new users and ensure proper logging of significant events.
If debug.log had printed a trivial message that it had started it's shutdown process, I wouldn't have been pulling my hair out trying to figure out why killing it was doing nothing. (I was watching tail -f debug.log)
Related note
Adding to debug.log that the local blockchain is behind by more than ~2 blocks and that some time may be required to catch up would also be very useful, especially for people unfamiliar with bitcoin[d].
