Why is it so complicated to use a second PC to plot/farm Chia?

I don’t know who from Chia might care, but this feedback is coming from someone who is not computer illiterate. I won’t call myself and expert, but I can figure anything out when it comes to computing.

The big issue for Chia right now is that last year things were booming. Now? There have been many new versions of Chia. So anyone trying to plot now, or use a second PC will be stuck with outdated guides.

I simply want to use 2 or more computers to plot Chia. To actually accomplish this task in July 2022, it’s a web of old crud online, outdated tutorials and oddball terminology. I really don’t think this could be explained worse by Chia on their own pages. Have they kept up with the new Chia versions and how that might affect guides or tutorials?

On this forum I don’t expect anyone to be able to relate. Most everyone has plotted out or has next to no need for a second anything. There are NO NEW ENTRANTS to point out some of the warts. I’ve been around since last year but only got serious about Chia lately. But as it is, people coming into Chia now or in the future are going to be pulling their hair out.

I suppose if Chia wants to remain fringe then leave it the way it is. Make it complex. Don’t simplify it. Leave it as is. It’s a terrible end user experience. You want to go big? Don’t make it this complicated.

I’m a simple man. I just want to use a f’ing second computer to plot with.

Harvester? Is that what it is? The instructions for setting up a harvester are enough to give anyone a headache. Is a second computer that I want to plot from called a harvester? Most videos don’t mention harvester when they show how to use a second computer to plot with. Maybe somebody should take the term “harvester” behind the barn and shoot it.

This should be as simple as installing the Chia GUI on a second computer and running it. Or having a Chia installation that is specific to a secondary PC that you want to use just for plotting or farming. As I found out, there are roadblocks right upon installation. I hit a bug. What about the Chia passphrase and the second computer? No idea! Do I even need to open the Chia GUI? I would actually have no idea. Tutorials show people opening up to the Chia GUI splash page to verify their keys. How does the Chia passphrase complicate matters? I literally have no idea. Go ahead and read the official Chia page on the passphrase and you will scratch you head and ask yourself WTF is this talking about. It simply adds an entire level of complexity.

Then of course nodes. Maybe some people understand why or how more than one node is bad. Maybe some people knows what happens to Chia if you open up the Chia GUI on 2 different computers on the same network.

So explain this. How do you navigate having to install Chia on the second computer, yet somehow avoid it turning into a 2nd node on the same network?

And yes, what about the passphrase? No videos showing that wrinkle in the process.

To summarize? This is like seeing your destination across the street. It’s right there. Should be as easy a F to get there. It’s just across the street. Why the F should it be complicated? It’s right there! But alas, Chia makes the process of expanding beyond one computer a F joke, to be perfectly honest.

So if Chia wants farming to be a club for elite computer nerds, great. The way it is now, for what I’m trying to accomplish, I’m not sure who could manage it. It’s odd right? They make the GUI in such a way that it’s simple to understand. Sort of. Yet outside of that box, they completely forgot about regular people who want to enter the Chia farming space.

I realize too why I put this off. I was reading and watching, trying to make sure I understood what I needed to do. Of course when I figured I’m good to go? Things like passphrases complicate matters to the point where I can’t get to where I need to go.

Go ahead. Read the Chia documentation on running a harvester. It’s hilarious actually. When I’ve seen that in the past I said to myself, no, that’s not what I’m trying to do. I’m just trying to run the Chia f’ing GUI (which I can’t do on a second computer) in order to plot on a second computer and fill hard drives. Far too complicated. But who can appreciate what I’m saying here? Likely nobody here unfortunately. And unless Chia price bounces, I’m sure this feedback will be buried and good for nothing.

In closing, nothing is really worse in computing than having a software that updates features, but the community and supporting tutorials don’t keep pace. You end up with 97% dog S solutions that are the only thing you can find when you need guidance. The web is full of clutter of outdated, useless information. I’m afraid Chia is in this place now. They wouldn’t know it because no new entrants. So I’m saying it here. Ultimately Chia needs to have the best and updated guides. For their own sake. But clearly they don’t know how to write for the farming audience. Maybe their guides are great for the elite computer nerds, but is that really their goal? I found the same issues when it came to updating the database. A lot of misinformation that made the process very painful. I wish this forum had new entrants so some of the questions can get updated answers. That is not the case however.

Oh well. Super annoyed if you couldn’t tell.

1 Like

I know the basics. And I might be able to help with some other items, too.
My help is based on solo farming.

Pool farming probably needs only some other key, but I am not sure what it is or how to configure it.

First, the layout:

You have your full node, which runs everything: farmer, harvester, and wallet.
As best I understand it, the farmer and the wallet should run only on the full node.
The harvester can run on additional computers. The harvesters send their results to the farmer that is on the full node. So only one farmer.

– Plotting on multiple computers:
This one is the easiest, because you do not need your mnemonics.
You need only your “Farmer public key” and your “Pool public key” (and if you are in a pool, then that would be some other key – which is confusing for me, because I farm solo and have to specify a pool key for my plots).

If you need to look them up, run the following on your full node (where you are running the GUI).
$ chia keys show

It will show you other information, but you do not need it for plotting elsewhere (unless a pool is involved, then you might need that key, too (I don’t know about pooling)).

To create plots, you run:
$ chia plots create -t [temp directory] -d [final directory for completed plot] -f [farmer key] -p [pool key]

There are other options that you can apply to the line, to specify memory usage, K size, number of plots to create, etc.

chia plots create --help (two dashes preceding “help”), to see a full list of options.

madmax is similar. You just need to locate its executable and basically do the same thing.

Your additional plotters do not need to be networked. They do not need to be on the internet. As long as you installed Chia, they will have the executable for you to plot.

To run additional harvesters, you need to tell the additional plotters the name (or address) of the location of your farmer, and will have to run “start_harvester” or something like that. Someone else will have to guide you for this step. I once read that it is simple, once you know where to enter the information.

If you start up the GUI on a second box, then it will start a full node, and that is bad.
Perhaps config.yaml can be configured to not start all of the services, and only the harvester?

The farmer is on the machine that your router forwards challenges to. So you can have only one farmer (unless you have more than one IP address). Your farmer will know where your other harvesters are, and will forward the challenges to them, and wait for the harvesters to reply.

By the way, I found the documentation lacking, too. That was roughly 12 months ago, and so I stopped checking. The folks in this forum were my documentation for getting me up and running.

It would be helpful if, when installing Chia, it asked a few questions, such as whether you want to run a full node or a harvester, and set you up accordingly. And it should offer help on what a full node is and what a harvester is, and what a farmer is, so that you will know how to answer.

Personally, I don’t think it’s difficult at all, but the Chia software really ought to have another ‘mode’ added. In the original releases, it was full node mode and that was it. You had to quit the full node and manually configure a harvester using CLI.

Then along came ‘wallet mode’ so you could chose that or full node mode.

Why not add a ‘harvester’ mode ? Select this option and it could walk you through configuring the machine as a harvester.

Anything is possible!


Thanks for your thoughtful reply and not blasting me.

Just a quick update. Of course after posting, I was able to get past some of the roadblocks. Some.

I appreciate your explanations of how all these parts go together. It is helping clearing the fog in my brain.

I admit I’m stubborn. I’m somewhat set on using the Chia GUI for plotting, although I understand and could use the CLI method. I realize too that everyone says you need to use command line. I say? I say that Chia should not be a command line based system. Sure for elites it’s fine. I could do it to. But why use a command line when to me it’s such a big step backwards in terms of usability. Remember games on computers used to comprise of just text? Now games use graphics. Command line in 2022 is a baffling direction. Hello Chia?

My experience so far can be called one word. Clunky. It doesn’t make any sense to install Chia on a system, but you actually shouldn’t run it. Really? What if I do? What breaks? Shouldn’t you be able to run Chia on a second computer on your home network without needing a computer programming degree?

I will absolutely be sharing my experience with Chia on this. I will gladly tell them what they need. Obviously they can’t see the forest for the trees. That’s fine. It can be explained.

Obviously Chia put a lot of work into the Chia GUI and design. It makes no sense to parse this out on other computers and require a 1980’s style command line solution.

You should be able to run a stripped down Chia GUI on other computers that you want to use for farming or plotting. An export function that will allow you to take those keys, and have them installed on the other computer without a bunch of nonsense. You should be able to take the vital parts from your original installation, and simply drop those (install) onto another computer on your network.

There is a slight problem I’ve encountered because I went unconventional, through the GUI. Entered the mnemonic phrase to get me in. I got in, but of course no database. It was an unworkable skeleton. The pool dropdown has nothing to choose. Yes, I’m guessing I could copy my keys from my main PC and enter those manually into the GUI and accomplish the same thing. But again, why so complex? Now I’m copying a backup on my database onto the 2nd PC.

And yes, I will need to start the Chia GUI on the 2nd computer with internet connected. Once loaded? Unplug internet and plot away via GUI. That is so clunky it’s baffling. Sure, sure. I’m not doing it right. I am suppose to not load up Chia. I’m supposed to delve into copy/paste in a command line. I guess I’m trying to prove a point of how ridiculous this is.

I will update as I get through this.

I will say though, at this stage, it seems common sense to have a different Chia GUI that can be used exclusively on other computers. Like a bedroom computer where you want to throw in a couple hard drives to add to your farm. Done in a user friendly way not requiring command line garbage. An export would work. Something that takes the required information to enable a simple installation on another computer. Period.

Get past that. An interface is the ideal. Is it not? To me a command line is required when all else fails or if you are some type of fringe software project. The opposite of mainstream and user friendly.

I think you mean that it should not be limited to the command line?

No farming on other computers. Only on the full node.

Ideally, the GUI should guide you, to make sure you do not double-up on farmers.

The GUI should also make it simple to run as only a plotter and / or harvester.

I do not believe that you need anything on the “harvester only” or “plotting only” PCs, other than the Chia application.

Same thing.
Unless I am misunderstanding something, your additional computers do not need an internet connection. Or are you saying that you need the internet (on the additional PCs) for activities unrelated to Chia?

If it is the latter, then you will need to kill the farmer process (or find a way to keep it from starting).

I solo farm and use Bladebit on a second machine for plotting with no difficulty whatsoever. I am using a Dell r910 with 1024GB of ram running Arch Linux. Bladebit requires 416GB Ram to run, so I run two instances of Bladebit when plotting. You have to work from the command line. To plot you would use a command like this: ./bladebit -f Farmer Public Key -p Pool Public Key /mnt/disk1
The CLI (Command Line Interface) is very useful for tasks that cannot be easily achieved with a GUI. I like using a GUI, but using the CLI means I have a far more customizable experience when using computers.

My GUI (1.33) offers BladeBit from its drop-down menu.

You got me! I did not know Bladebit was on the GUI because I have been using Bladebit from the CLI since Sept 2021. It just worked, so I never bothered to change. I don’t use the Chia GUI for farming or plotting because I prefer Machinaris and Bladebit from the CLI. I only use the Chia GUI for my wallet.
Why do I favour the other Chia products? I found the Chia GUI a pain to use. If you are plotting, you do not need a synced wallet, it is a big overhead.
So, I have Machinaris running on my 24/7 server with all the other things I run, like Pihole, HomeAssistant, etc. I run the r910 like a power tool when I need to plot and switch it off afterwards because it consumes much power (500-1100W) but can plot 32 TB in 24 hours.
So if you are saying the Chia UI is funky and poorly documented, then I would agree. However, being a nerd, these limitations provide an excuse to tinker.


I plot via the command line. But I run the GUI, because it starts up all of the services.
I know that the services can be started via commands. But when I was new to Chia, the commands were not yet known to me, and since the GUI just worked, I still use it (but not for plotting).

Also, the GUI gives me real-time XCH info, and real-time Sync status.
I find it useful as a monitoring tool, but not as a plotting tool.

1 Like

Farming on many machines · Chia-Network/chia-blockchain Wiki (github.com)
This should help, I have one main Farmer and three harvesters and easy to setup with that link.

Actually, this is strikingly simple.

  1. Forward port 8444 to the primary machine’s (doesn’t matter which) IP address on your router.

  2. Run the GUI on the 1st machine (add ur key).

  3. Run the GUI on the 2nd machine (add ur key).

  4. Repeat on others.

Plot and farm away on them… even on the same network.
End of procedure.

May not be as recommended, nevertheless works. Once you get acclimated, you can change it to done as Chia (tries to) tell you how.


I think this is exactly what @MisterSavage is talking about. Every time I have looked at this tutorial, I went “well never mind then” I’m sure I could get it to work, but it’s anything but simple.

Problem is that a lot of people here are so used to the fact the fact that they understand all this kind of stuff that they forget how many people will get stuck after line 3 or 4.

It’s simple, you just use this and then and then set port this to that, setup a link between bla and bla, run this command, set these parameter, shields to full, inertial dampeners to max and away we go… I help people in our office regularly with things like print jobs being stuck in the cue or “my hard drive is full, what should I do?”. Some of these people make more than me…
Port forwarding on a router? Most people call their ISP tech support for that.

GUI is not a term in common usage, neither is CLI. It’s called an app, you click on it and it will tell you what to do.

1 Like

I would agree with the call outs, I’m pretty good with a PC and technology and some guidance and instructions, but I have very little knowledge of command line prompts and how to configure and view normal PC related items. Couple that with Linux being preferred by many and trying to keep straight what can be done in Windows and what can be done in Linux it gets even worse.

I took a break from chia after getting started last year and came back this year trying to get back into it and have been spinning trying to figure out how to setup a bladebit plotter on a second machine, how to best connect all the drives so he read times were sufficient, etc. I’ve got an external 64 bay enclosure connected to the full node over sas, but also have a SATA card plugged in wiht several drives, and then also several USB drives. I know this is not ideal, but not really sure how to break it up on several machines to make it efficient.

Would definitely be nice to have simple options in the GUI that says, full node machine, harvester setup, plotting only, etc. Statistics on read times, I’ve slowly started to find 3rd parties that will give you some of this info that will be trying shortly. I have been successful with 3rd party GUI for madmax on windows without the chia GUI, but even that was a bit tricky until i realized I just needed the chia files on the machine, etc.

I will say the newest GUI gives you way more info than the previous, but if your trying to go with a large operation you really need to be a IT specialist, software Dev, or equivalent that is comfortable and familiar with command line prompts and interacting with the backend data.

Don’t even get me started on the NFTs and CATs….no clue what that is just yet, haven’t gotten that far in my research to know what those are just yet lol.

1 Like

Your right, its not easy to work this stuff out, I’ve being do plots regularly since I started in May 2021, but I’ve just got the cmd line saved, so I just paste it in and alter the number plots, so I’ve just had to work out where to get relevant info from.

You really don’t need to install Chia on your plotting computer. I’m going to presume you are on Windows. on the main PC which has the GUI installed do the following.

  1. Open file explorer and type %localappdata% in the address bar and hit enter
  2. Double click each of the following folders
    app-1.4.0 (note the 1.4.0 changes with each new installed version)

At this point if you click in the address bar and type cmd and press enter, it will open a cmd window and you can run Chia commands. You need to type chia keys show and press enter.
Find the farmer public key, double click the long number/letter combination so its high lighted, then right click the top of the window, click edit, click copy. Create a text file on a USB stick, type Farmer public key -f paste copied text

Now go back to file explorer and in the daemon folder you will find two folders Bladebit and Madmax, copy the contents of each of those folders to your USB stick.

To create portable NFT plots you will need the pool contract address - this is misleading, as you need this even if you are not using a pool, but you are solo pooling. To get this go to the pooling tab in the Chia GUI./APP, on a fresh install I think this is blank and you need to add a Plot NFT, so do this if its blank. Once you’ve got plot NFT showing (you only need one - don’t create more of them) it will have a name top left with a little ? mark next to it - hover over the question mark, and then copy the pool contract address. Paste it into a text file on your USB stick and put a reference next to it Pool contract address -c paste copied text

So on your usb stick you should have a text file containing your farmer public key, and your pool contract address, and you should also have the bladebit.exe, chia_plot.exe and chia_plot_k34.exe files. Those last two are madmax, and the first does only K32, and the last K33 and K34.

Now copy the files off the USB stick onto your plotter PC, say into a folder called plotter. Open a CMD window - use the method above using explorer but navigate to the folder you’ve just created.

I’m going to presume you’ve setup your temporary locations, and you’ll have to edit the following line to reflect those locations, number of threads to use ( the -t and -2 are the temporary folder locations, -d is destination, -r is number of threads) and your keys

chia_plot -k 32 -x 8444 -n 1 -r 36 -K 2 -u 256 -v 128 -t I:\ -2 R:\ -d D:\ -c YOUR POOL CONTRACT ADDRESS -f YOUR PUBLIC FARMER KEY

For Mad Max you can lookup what the switches (-n, or -t etc) mean at GitHub - madMAx43v3r/chia-plotter

Using Bladebit is very similar, run bladebit -h from the cmd line for explanation of the switches.

I built myself a little database so I just enter the relevant details (K size, number of plots) and it then coppies the complete cmd line to the clipboard and I can then just paste it in, hence why I forgot most the above and had to google some it to find out where to get the keys.

Oh, and the reason you only want to create one Pool NFT is because, A) Once created you can’t get rid of them, and B) plots created on one pool NFT can’t be used on another. Only useful if you wanted to split your plots between different pools or self pool and pool some.

Hope that helps, and yes it really should be easier.

I have everything on one fairly powerful Windows computer (Full node, plotting, farming harvesting etc.) The following might be useful info for someone.

At the moment I am plotting k34 (sometimes k33). I have a 1TB SSD (F:) and a 480GB RAM Drive (R:)

I use the MadMax plotter. If I want to do a single plot, it seems to make no difference in whether I plot from CLI or the GUI.
For multiple plots I use the CLI because it MadMax has the nice feature in that it will start the next plot at the point it starts the copy of the previous plot. What you have to watch out for here is to hope that the copy is fast enough to complete before Phase 1 of the second plot writes too much data. In my case it’s OK.

So these are the command lines I use to create a plot for SpacePool:

Open Windows Terminal (Admin) - not sure if you need Admin mode but that’s what I use.

cd c:
cd \Users\graha\AppData\Local\chia-blockchain\app-1.4.0\resources\app.asar.unpacked\daemon\madmax

.\chia_plot_k34 -k 34 -n 17 -r 8 -t F:\Chia\ -2 R:\Plot-RAM\ -d \Synology-1821\Seagate-13-8TB\chiaplot\ -c xch20edyfnk4d67k278mkp40tzln8e7c5906ys0w0w22tr2sk3s678nnq27erh7 -f 8g9kl00d06ac907cfc3s5dv66jk88423189a1db39a6737hkj944f9a157c4e69d074fc99efdf096732a54ff1daws35fvgg7h

To plot k33, just change -k 34 to -k 33
-n number of plots
-r number of threads
-t temp address, in my case the 1TB SSD (F:)
-2 temp 2 address, RAM Drive (R:)
-d destination folder, don’t forget the final backslash (certainly for Windows)
-c SpacePool contract key
-f Farmer public key

I appreciate all the replies and help sorting through this. I’m glad some here have walked a similar walk. It’s great to have a LOT of information in one place. Yes, running on Windows 10.

What I really find with computers is that when you have complicated software or processes, it obviously involves 10’s or 100’s of steps. There is no doubt that when you put this on the shelf for a month or two, you almost have to start from scratch to remember all the hoops that you must jump through.

Because I’ve set out on a path for using the Chia GUI, I need to see this through. For really my own sake. Following this, I will likely stick with the command line processes which has been outlined here by members.

So if someone out there says to themselves, they want to plot on another computer on the same network and prefer sticking with the Chia GUI which has three different plotters included which will avoid having to learn about command lines for Mad Max or Bladebit, I’m going to explain.

Of course this is unnatural. I would call this more of a workaround rather than a method. Anything this clunky is not designed for use.

I copied over the database from my main computer. This meant the second computer had what it needed so the GUI had my plot NFT in the select your pool section of the GUI. Without that database there was obviously (not really that obvious until you try) no plot NFT to plot to. If you didn’t do this and just plotted, those would be…OG plots.

As for the limitations of using Mad Max in the GUI, if there are settings that the CLI can do but the GUI can’t, then I would suggest Chia needs to roll out those fixes or improvements. If you can do everything from within the GUI, I think that should be their goal. It shouldn’t be good enough that people need to go outside of the GUI because that’s the only way they can do certain things. The GUI should cover all the basis. If not, that’s a failure in my books. But I digress…

So back to this. The clunky is this. You need to start your computer connected to the internet. Start the Chia GUI. Then disconnect from the internet, which stops the GUI from trying to sync. At that point, the second computer can be used as a Chia GUI plotter.

This is where it’s at for me right now. As I mentioned, because I had the goal of using the Chia GUI for plotting on another computer on my network, I had to see this through. I’m not sure I’m exactly there yet. Why would I bother you ask? Well I can tell you that at least one other person was asking about using the Chia GUI for plotting on a second computer. The answer provided was “yes”. My mistake I suppose was seeing that “yes” as meaning, sure, no problem, no roadblocks, no hassle. WRONG.

The way I see it? If the Chia GUI isn’t the best option for Chia, then they need to fix something. If there are plotting shortcoming in the Chia GUI? Those should be addressed. If I’m involved in the Chia project and I read that the Mad Max plotter in the GUI is missing some settings, thus making CLI the must-use solution? That is something I would be fixing. Nothing should be better than the Chia GUI. It’s your freaking showcase! It’s what any new user uses. If it is deficient in any way? Fix the damn thing.

A lot of smart people at Chia. If you cannot simply use the Chia GUI for plotting on a second computer? Do tell why that isn’t something you want to fix. Is that because no new entrants into Chia farming who can voice their displeasure? If so, then please take my feedback as legit.

Chia needs a GUI shell of sorts, that contains all the information that a second computer needs in order to plot. An export functionality of sorts. A simple GUI question of, is this a second computer on your network, do you want to plot on this computer, etc. As already mentioned by someone else in this thread. Software is suppose to make things easier and faster for the end user. Manually plotting seems so backassward to me. If manually entering settings is the only way to accomplish certain plotting requirements then Chia should have already seen these shortcomings and fixed them in the GUI so that there isn’t a need for manual, outside of GUI plotting. People can, sure, but there should be no benefit to it. The GUI should be the go-to solution. Period. But clearly it cannot be in its current form. But it should be and could be fixed.

You’re lucky you don’t have to use programs like Swar Plot Manager, that would of really fried you’re brains. My first plots took around 9 to 12 hours, using Swar plot manager would plot multiple plots at the same time but staggered, all taking 9 to 12 hours, and if it crashed or Windows decided to do an update you lost the lot that was in progress. We really have moved on a lot, but it’s still not simple enough.

Farmers are just a means to an end, what’s important to Chia is use cases for the XCH blockchain, we farmers come second to that, always will. The fact there is 22EiB means there is enough farmers, I think 5EiB is enough net space.

I know that NFT stands for “Non-fungible token”, but I have no clue how to create the token, or what they are used for, etc. For CATs, I do not even know what that acronym means.

Funny, but yes, I was set to use the SWAR plot manager in fact. Recently. Then I discovered a bit of a trick within the Chia GUI that was a lot easier.

On an old PC with i7 4th generation, 8 threads and 16gb of DDR3 ram using spinning disks I get a plot done in about 3.7 hours. I can set up the jobs in the GUI, the right amount to set it and forget it. Come back and the drive is filled. Zero need for manual anything.

I can’t seem to figure out MM at this point with my hardware. What I did see though was my CPU being pounded at 100%. I don’t care about a lot of things but I care about my CPUs. I still haven’t number crunched every last detail yet in terms of power costs vs. buying SSDs to plot with vs. wear and tear on a CPU. I’ll eat a bit of added power if that means I’m not killing hardware. Not looking to buy extra CPU heatsink for the purposes of shaving 1 hour off my plot times. But still a work in progress. But I digress.

Once this second computer runs for a bit, I’ll add a third plotter to this mess of clunky. Actually I don’t think it should be too bad. I’m excited to compare my power consumption vs. hardware costs vs. plots per day. I don’t believe in new hardware for farming/plotting Chia. Sometimes you must lead by example and face backlash from those who deem popular methods or mainstream methods as the “proper” way of doing things. Sorry, I’m trying to prove otherwise.

So long as your CPU is properly cooled your not going to kill your CPU - I’m more concerned about the hard drives over the next couple of days - supposed to be getting up to 35c here on Tuesday.