Closed Himan2001 closed 6 years ago
I expect this problem will resolve itself when the load problem is resolved. Refer to #44
I can confirm this problem on my linux-machine: The load is way to high and the logfile shows enormous dropped network connections to the connected clients - this makes continous syncing unstable and unreliable. I see that working nodes got as "tryed" connections to the peers.dat db and the daemon search for other peers. Without limiting the maxconnections=15 i noted 1 time in my NAT-router a NAT-table overflow a few days ago. The Router log showed more then 300 connection-attempts in a very short time, so the firewall considered this as an internal attack and blocked the communications.
Acording to the usual sync-problems on Linux and Windows i have seen, that the Sync hangs mostly on a Blockchain that is generally up-to-date. The Sync is starting again often only when a NEW block comes in. By the actual slow blocktime it results in a few minutes delay to get in sync again.
I monitored Linux and Windows with the same config/peers and blockchain hight directly after startup and booth daemons are hanging nearly 8 Minutes. When a new valid block comes in, and booth systems was in sync in only 2 seconds.
By the way, at the time of "waiting for nothing" the load is always nearly 100%