I am trying to connect GPU farm to Nicehash, but there are problems with frequent forced disconnects every few minutes. All machines are connected through stratum proxy (Stratehm). There are extranonce subscriptions enabled both in proxy and miners and notification is working properly.
Similar problems appear when machines are directly connected without proxy - connections are dropped in several minutes - again extranonce subscription is checked on Nicehash site.
It seems like there is something closing the stratum connections even if extranonce subscription is active and working correctly. I saw this in a post above:
>>>Unfortunatelly, we are under constant attack by botnet miners, therefore we have filters in place that ban IPs that only idle. If you have large hashing power and static IPs, contact us for a whitelist.
Is there someone else with many machines experiencing the same problems?
we had the same issues ... with a very similar setup ...
we use stratehm 1.5 ( not the latest version ) and our farm worked perfectly until we were booted for idle ...
nicehash were very supportive of this and have suggested a solution ... we still get kicked on occasion - but that is normal and it reconnects pretty much right away ...
the reason is that their stratum DOES ban the ip that you are connecting the stratehm proxy from ...
the only way is to send support a message and get that ip you are running from whitelisted ...
nicehash have been very helpful in this area - and really have had no issues since the whitelist ...
we have requested the dev donation server ips (
https://bitcointalksearch.org/topic/a-chainworks-industries-cwi-project-donate-by-mining-dbm-1089744 ) to be whitelisted also - and nicehash have also taken those into the whitelist within 24hours of request ... these servers will be active and fumctional by the end of the week when the ips of the new servers get whitelisted also ...
have had nothing but excellent help and service from nicehash support ... quite refreshing compared to other pools / hash brokers that we were looking at as the main sites to allocate all the hash to ...
cant ask much more than that ...
#crysx