[x] discoverTask starts, regardless of number of concurrently running tasks. The task is started as soon as it's generated, instead of being added to a queue. The discoverTask generating conditions remain the same as that of the original geth logic.
Discovery happens much more frequently now, but it still doesn't seem to occur as often as normal geth.
Maybe force discovery every 4 sec even if the currently running hasn't finished?
[x] DialCheckFreq flag. Default is 15 seconds.
[x] AddBlacklist JS console function currently allows updating TCP dialer/listener's setting dynamically. The UDP dialer/listener's setting should be updated as well.
[x] Reverted TCP dialer with cancel context back to normal dialer.
Running tasks are ignored during p2p server shutdown. There is no need for manually cancelling them.
[x] Reverted dialTask behavior back to original.
We DO want node-finder to resolve address of static nodes if it fails to connect to them.
[x] New log level TASK for New dial task and Dial task done messages.
[x] New log level PEER for messages related adding or removing p2p/ethereum peer connections.
discoverTask
starts, regardless of number of concurrently running tasks. The task is started as soon as it's generated, instead of being added to a queue. ThediscoverTask
generating conditions remain the same as that of the original geth logic.DialCheckFreq
flag. Default is 15 seconds.dialTask
behavior back to original.TASK
forNew dial task
andDial task done
messages.PEER
for messages related adding or removing p2p/ethereum peer connections.