Instead of catching and logging errors/ panics, we can fail fast and panic, so that errors can not be missed. They would crash the node.
This is not intended for production setup though.
One solution proposed was a compile-time flag. Other options are start parameter or environment variables. Other ideas?
Follow up on #32
Instead of catching and logging errors/ panics, we can fail fast and panic, so that errors can not be missed. They would crash the node. This is not intended for production setup though.
One solution proposed was a compile-time flag. Other options are start parameter or environment variables. Other ideas?