This commit makes the rebroadcast logic being run in a goroutine. This
is done to ensure we don't block the whole broadcast loop while
performing a rebroadcast.
Rebroadcasts can be slow if the number of transactions are large (each
broadcast has a timeout of 0.5s), which would cause publishing new
transactions to block. With this commit we let the rebroadcast continue
in the background, while we can keep pushing new txs.
Coverage increased (+0.09%) to 68.43% when pulling 6f292a7e2f35c763acb2609116c4f9f22a37553b on halseth:push-tx-async-rebroadcast into a655679fe131a5d1b4417872cc834fc3862ac70e on lightninglabs:master.
This commit makes the rebroadcast logic being run in a goroutine. This is done to ensure we don't block the whole broadcast loop while performing a rebroadcast.
Rebroadcasts can be slow if the number of transactions are large (each broadcast has a timeout of 0.5s), which would cause publishing new transactions to block. With this commit we let the rebroadcast continue in the background, while we can keep pushing new txs.
TODO