Closed matthewdarwin closed 1 year ago
hi @matthewdarwin would you share your exact commit hash, OS, config, and the steps to re-produce? is this random behavior?
I believe the step to reproduce is to keep running curl on /v1/chain/get_info while shutting down nodeos. It has been seen on many versions going back to at least 2.0, so not specific to any version.
Will track this in linked leap
issue now
Using: mandel 3.1-rc3.
Notice 8 minute delay on shutdown.
The "corrupted size vs. prev_size" is not a new problem (happened in earlier versions of nodeos). would be good to understand what is happening though. Seems to not cause any problems as nodeos starts fine next time from the written state.