Farm too large to fit on a single node

My farm is too large to fit on a single node, so im wondering about the options for splitting it into 2 nodes with 1/2 the plots on each node. Can I use the same “key” for both GUI’s if they have different plots or do I have to use a new wallet and new key for the second node?

What does that mean? How do you know it’s too large?

I can’t fit any more storage on my node. Ive used all the available space. 2x 16TB disks and I have 28.7TB used, so it’s almost 100% full.

Check out the remote harvester option: Farming on many machines · Chia-Network/chia-blockchain Wiki · GitHub

Hmm, not sure about those details, looks fairly complex. It seems to be working just having the 2 nodes running.

Multiple nodes is problematic, because you can only have 1 node on your network publicly accessible on port 8444.

That’s nowhere close to 100% full.

You can put more drives in external drive housing and connect with USB too.

Yes, but that is a zero-sum game. No matter what I add, ill fill it up and be in the same boat. The only real solution is horizontal scaling to new machines where I can scale indefinitely.

Running multiple harvesters is the intended solution for scaling horizontally.

1 Like

Yes, I saw the reference to running multiple harvesters, but I haven’t yet figured out the distinction between a full node, a harvester, and a farmer. Currently, I only know about the “GUI”, which appears to be all functions rolled into one unit. I just need to get a more practical understanding of what it means to run a harvester as opposed to a GUI node. As of this time, the simplest approach is to run multiple GUI’s, one on each machine. It seems to work as I was able to win a block with this configuration, so just wondering why the need to build a harvester when the GUI seems to do it already.

Did you read anything on the Wiki? It’s explained there what the difference are.

Of course I did, but that doesn’t mean it all makes sense.

From the Beginners Guide:

The Chia software is made up of several parts called daemons. Each daemon does different things. The main ones important to beginners are:

  • node - This syncs your computer with the blockchain
  • farmer - The farmer sends out proof requests to the harvester
  • harvester - The harvester checks your plots to see if they satisfy the proof. If you have the best proof, you win coins.

Assuming that your machines are behind a router and are NAT’d (i.e. they do not all have there own IP address on the internet) then only one of your full nodes will be able to receive connect request from other nodes. So one will have many connections (40-60 based on conf settings) and the others will have fewer connections since all their connections will have to be originated by them (no incoming to accept). I’ve seen cases where it can take a while for a node to find a connection when it is only originating connections. I don’t understand the bigger picture networking design of Chia to know what the implications of having fewer connections might mean to your farming. Maybe someone who has studied this can chime in.

1 Like

Ive been reading and trying to figure out the details of setting up a harvester, but I tried it and I think a few steps may be missing. I couldn’t get it working. This was on a VM just for testing purposes. Im going to re-try using a physical machine instead of a VM and see if that helps. May take a day or two, in the meantime, it does appear that each of the full nodes is getting “Plots Passed Filter” numbers above zero, so im assuming it’s working. In addition, im up to 4 xch, so I know I am getting awards.