Nicehashminer: CPU miner interupts - XSOCKET ERROR - RECEIVE error: socket closed
CPU miner interupts using xmr-stak-cpu. Cpu cryptonight results on nicehash web does't look continuous, it gets interupted every 5-10 minutes. I got the same results by testing separatly few of my workers with NHML and NHM:
Don't know if it's related but I notices that I get this error every 15-60 minutes (not as often as interupts on the web) on all 50 workers.
It reconnects in 10 seconds and repeats after some time.
Tried to change api ports - the same.
Tried looking into the miner configs, at least reduce reconection timeout, but this led to another bug: https://github.com/nicehash/NiceHashMinerLegacy/issues/112
utc99
All 6 comments
Same here. Maybe server inconsistencies?
ChriscomIT
on 3 Aug 2017
ditto for me... "socket closed" thing cost me double digit percentage block rejection, in the end i stopped mining xmr altogether.
i have tried pointing to various server location, but no dice.
lmlim
on 4 Aug 2017
This is because Cryptonight takes a long time to submit shares (> 10 min sometimes), especially when mining on a CPU. The NiceHash display has gaps because during those periods no shares were submitted, but be assured that you are getting paid for the mining you've done during that period (once the share is submitted, you can see the spikes).
As for the socket disconnections, this is because NH will timeout the connection if a share isn't received in 10 min. Since CN on CPU can take longer than this to submit the share, the connection timeouts. The miner reconnects in 10 seconds though so there is not much lost.
DillonN
on 4 Aug 2017
@DillonN Thanks for the detailed clarification! The pps explanation sounds logic. Much appreciate it!
ChriscomIT
on 4 Aug 2017
Perfect explanaition, so it's an expected behavior and no need to worry about it, thanks. Issue closed.
utc99
on 4 Aug 2017
Given the explanation above, surely it should be a 'warning' rather than an 'error'.
twocs
on 20 Feb 2018
Related issues
tham2539
路
4Comments
LuckyGeniusss
路
4Comments
ghost
路
4Comments
liranscu
路
5Comments
Kayfolom
路
3Comments
Most helpful comment
This is because Cryptonight takes a long time to submit shares (> 10 min sometimes), especially when mining on a CPU. The NiceHash display has gaps because during those periods no shares were submitted, but be assured that you are getting paid for the mining you've done during that period (once the share is submitted, you can see the spikes).
As for the socket disconnections, this is because NH will timeout the connection if a share isn't received in 10 min. Since CN on CPU can take longer than this to submit the share, the connection timeouts. The miner reconnects in 10 seconds though so there is not much lost.