…would it be enough? I have my server in remote location and there is 4g/lte connection only - currently using one but it is expensive. There is however one isp which gives unlimited 4g connection in 128kbit/s speed - would it be enough?
I think it should be because 128kbit/s is around 1.3GB for 24 hours but so far, chia is “eating” around 0.6GB per 24 hours but I will better ask - anybody with experience like this? 128kbit/s just for chia linux machine…?
I don’t want to be that guy that always says flexfarmer!
Just checked mine, I’m using apx
280 up
400,000 down
Guessing you would struggle
mind to share also units? have no idea what are those numbers 
KiB per ?
They’re listed on farm tab in the expandable window at bottom, but it gives no more info than that.
My full node eats about 3.8 GB downloaded and 1.5 GB uploaded per week. This is fully synced, on 24/7. If you need to sync, it will be much more than that.
Next problem you may face is stales. Full node is very sensitive to the speed of your connection. I used to have a wi-fi connection with default settings and it got me a lot of stales. After changing to “high performance mode” in power settings, the stales went away.
Next possible constraint is number of peer connections. Chia protocol on network level is similar to Bittorrent. It requires lots of p2p. Some ISP detect it and impose limitations, thus hampering your operation,
To really test your setup, join a pool that keeps track of your stales. If you’re getting more than 1% stales, it is no good.
Alternative option is to use a harvester, connected to a full node hosted in a cloud (that has your farmer secret key). Flexpool adopted this model: they have a full node deployed on their hosting and you install a proprietary harvester “Flexfarmer”. It will consume a lot less resources, essentially because it only connects to one peer.
this is intersting but I don’t see any stales in my log. looking with cat debug.log | grep "stales"
also that isp thing - I always have 8 nodes connected to mine - never had higher number - can it be related to some isp limit???
I don’t know how to detect stales in debug.log, but take a look at an example how Flexpool counts them. This is one of their miners listed as top 10:
Scroll down a little and you will see number of successful partials vs stale partials.
As for the number of peers, I believe the “normal” size is ~100, but that’s what I saw other members posted on this forum. Perhaps someone more competent can chip in.
Stales is a pool term. I believe they’re just slow responses.
So if your solo, and aren’t showing WARNING for your lookup times being over 5 sec you should be good.
haha. Soon there will be a reply on Flexfarmer
