zed-wong opened this issue 2 years ago · 0 comments
The current method would be slow when there are a lot of tx coming in. Should rewrite a multi-thread version.