Please give more information about Remote Compute

[mbegunok]

Please gove us more info about Remote Compute.

Scheme with additional compute machine. What should be set in CHIAPOS_RECOMPUTE_HOST= on both machines? Shoud we run chia_harvester, chia_farmer, chia_full_node on both machines? Should we run run_recompute_server on both machines?

[endurance1968]

on the compute just run the recompute server. nothing else needs to run there. on the client set the variable and start the harverster nothing else need to run there. of course you need the farmer as well - does not matter where you start it - as long as the harvester knows where.

[unsaloc]

"set the variable" which variable? :)

[mbegunok]

"set the variable" which variable? :)

CHIAPOS_RECOMPUTE_HOST . I got 3 different servers. 2 of them are currently making new C8 plots. I would like to make a third server work as a failover machine + farmer. Whatever I do, It sees only one server. Sometimes first, sometimes second

[endurance1968]

"set the variable" which variable? :)

as posted in the post directly before my answer so I thought this should be clear (as also mentioned in the documentation) :) I currently use two computes both are used.

starting the harvester on my linux with a scrript like:

export CHIAPOS_RECOMPUTE_HOST=<IPCOMPUTE_I>,<IPCOMPUTE_II>
/xyz/chia-gigahorse-farmer/chia.bin start harvester -r

[Dakes12]

Hello,

I need some help to run the remote compute on windows.

Just to be sure that I understood well: On the farmer (gpu equiped) I have just to start recompute server, without doing anymore settings? On the harvester I have to set the variable for CHIAPOS_RECOMPUTE_HOST variable via system settings to farmer ip, then to start harvester?

Thanks!

[Kruger1981]

@Dakes12 was about to ask the same and something clicked so I got it working I think.

Yeah on server side just start the remote compute server.

On harvester just create the variable and restart the farmer.

Variable Screenshot att.

Question, on server side I can see that the requests are being received, but at the end there is fail=0 I suppose its OK if there was a problem would appear fail=1, right? Can anyone confirm please?

Many thanks in advance and happy farming.

[Dakes12]

@Kruger1981 i have started the server on farmer, setup the CHIAPOS_RECOMPUTE_HOST to farmer's ip, started chia harvester. In server, i am not receiving any request as you have in the above photo. What i did wrong?

Thanks!

[mbegunok]

ok, it is really strange. Sometimes you need to use command ./chia.bin stop all -d and then use export CHIAPOS_RECOMPUTE_HOST=192.168.3.45,192.168.3.54 (whatever is your IP) and then use ./chia.bin start farmer on all computers. Sometimes you need to use this command only on the harvester computer, sometimes only on the farmer computer. Just try all options and wait a few seconds to look at the ./run_recompute_server.cmd screen. Results may be wary

[Dakes12]

I am checking it again. Btw, i am using Windows 11, in case that matters.

[Kruger1981]

Should be fine, i tested under w11 and worked.

On Sat, 12 Aug 2023, 17:06 Dakes12, @.***> wrote:

I am checking it again. Btw, i am using Windows 11, in case that matters.

— Reply to this email directly, view it on GitHub https://github.com/madMAx43v3r/chia-gigahorse/issues/173#issuecomment-1675980787, or unsubscribe https://github.com/notifications/unsubscribe-auth/AZGOI3VCZHASWNJB7YA6OWLXU6SXLANCNFSM6AAAAAA2VXJGI4 . You are receiving this because you were mentioned.Message ID: @.***>

[Kruger1981]

I think the export...xxxx is only for linux

On Sat, 12 Aug 2023, 17:03 mbegunok, @.***> wrote:

ok, it is really strange. Sometimes you need to use command ./chia.bin stop all -d and then use export CHIAPOS_RECOMPUTE_HOST=192.168.3.45,192.168.3.54 (whatever is your IP) and then use ./chia.bin start farmer on all computers. Sometimes you need to use this command only on the harvester computer, sometimes only on the farmer computer. Just try all options and wait a few seconds to look at the ./run_recompute_server.cmd screen. Results may be wary

— Reply to this email directly, view it on GitHub https://github.com/madMAx43v3r/chia-gigahorse/issues/173#issuecomment-1675978930, or unsubscribe https://github.com/notifications/unsubscribe-auth/AZGOI3WBZWFY4PDSBPY2ZP3XU6SOXANCNFSM6AAAAAA2VXJGI4 . You are receiving this because you were mentioned.Message ID: @.***>

[Kruger1981]

Did you namaged to get it to work?

On Sat, 12 Aug 2023, 16:54 Dakes12, @.***> wrote:

@Kruger1981 https://github.com/Kruger1981 i have started the server on farmer, setup the CHIAPOS_RECOMPUTE_HOST to farmer's ip, started chia harvester. In server, i am not receiving any request as you have in the above photo. What i did wrong?

Thanks!

— Reply to this email directly, view it on GitHub https://github.com/madMAx43v3r/chia-gigahorse/issues/173#issuecomment-1675973937, or unsubscribe https://github.com/notifications/unsubscribe-auth/AZGOI3WHCPGMFUQR2THLM33XU6RK5ANCNFSM6AAAAAA2VXJGI4 . You are receiving this because you were mentioned.Message ID: @.***>

[Dakes12]

Not yet. I will try again and i will get back. Thanks for support.

[Dakes12]

Hi, I tried again, without success... Just to be sure.

On the server side (GPU equiped), i started recompute_server and chia farmer. On the harvester, i have pointed chiapos_recompute_host to server local ip, then i started the harvester. This is right?

Or the chia farmer should run on each harvester and server runs only recompute_server?

Thanks!

[Kruger1981]

farmer only needs to run on 1 machine, regardless if it is on the same machine as the recompute server.

I have the following:

Chia gigahorse farmer running on: 192.168.50.180 Harvester 1: 192.168.50.185 Harvester 2: 192.168.50.44 Remote compute host: 192.168.50.191 and harvester 3 as well.

My farmer is capable of dealing with its load as it has its own dedicated GPU However Harvesters 1 and 2, are low spec machines and not capable of handling the load. So i have the system variable on these 2 harvesters pointing to 192.168.50.191 (recompute server)

When I 1st ran the recompute server, had to allow it through windows firewall, did you do that? maybe worth testing disable the firewall on the recompute server and on 1 harvester. Also remember that you can only start the harvester after you have configured the system variable on windows (as on the screenshot i sent before) Some say that after setting the variable you should reboot windows, but I didn't do that and it is working fine. I think your issue may be the firewall.

[Dakes12]

Thanks for answering.

Another question. Do i have to setup the wallet on each harvester, or it's enough just to start chia harvester and it will know where to farm? Remote compute host 192.168.50.191, is running any of farmer of harvester? How it's pointed to the farmer?

I haven't received any notification from firewall to allow recompute server.. Maybe that is the problem. I will check it.

[Kruger1981]

on harvester machines, I only run the harvester. On your farmer, do you see your harvesters? something like:

[Dakes12]

I am using only gigahorse farmer, without chia gui. I will give a try tomorrow with chia gui on.

Thanks!

[Dakes12]

I am still figuring problems.. Harvester is not sending any data to farmer’s recompute server .. Ports allowed on firewall, app allowed on firewall.

I have to follow chia official tutorial for using multiple harvester?

It seems that:

1. running server and gigahorse farmer on the farmer
2. Setting chiapos_recompute_host to farmer ip via system variables then running gigahorse harvester (restart included).

Still not working.

[Kruger1981]

Yes you need to setup the harvesters properly first.

Have you checked if your harvester shows on the darmer GUI?

When you setup the harvester did you ran chia init -c "whatever folder you placed a copy of the CA files"?

Seems to me that your harvester may not be setup correctly.

On Sun, 20 Aug 2023, 18:50 Dakes12, @.***> wrote:

I am still figuring problems.. Harvester is not sending any data to farmer’s recompute server .. Ports allowed on firewall, app allowed on firewall.

I have to follow chia official tutorial for using multiple harvester?

It seems that:

1. running server and gigahorse farmer on the farmer
2. Setting chiapos_recompute_host to farmer ip via system variables then running gigahorse harvester (restart included).

Still not working.

— Reply to this email directly, view it on GitHub https://github.com/madMAx43v3r/chia-gigahorse/issues/173#issuecomment-1685352335, or unsubscribe https://github.com/notifications/unsubscribe-auth/AZGOI3SOTOUA3DTPMOTABDDXWJE5XANCNFSM6AAAAAA2VXJGI4 . You are receiving this because you were mentioned.Message ID: @.***>

[Kruger1981]

Either later on or tomorrow I will be setting up a new harvester, do you want me to make a video?

[Dakes12]

Either later on or tomorrow I will be setting up a new harvester, do you want me to make a video?

Yes, please. Thank you!!!

[Dakes12]

Managed to see the harvester, but still no pings. In the log got DEBUG:

Firewall is off on both machines. imported ca ssl keys from farmer, run chia init then deleted temp ca.

[Kruger1981]

Did you imported the CA keys into the harvester?

On Sun, 20 Aug 2023, 20:48 Dakes12, @.***> wrote:

[image: image] https://user-images.githubusercontent.com/141576290/261873398-3ac9e90f-29fa-4160-a25b-1bebc76a4e6f.jpg Managed to see the harvester, but still no pings. In the log got DEBUG: [image: image] https://user-images.githubusercontent.com/141576290/261873487-b0e65070-5d97-469f-983e-d549cf86f524.jpg

Firewall is off on both machines.

— Reply to this email directly, view it on GitHub https://github.com/madMAx43v3r/chia-gigahorse/issues/173#issuecomment-1685375926, or unsubscribe https://github.com/notifications/unsubscribe-auth/AZGOI3SSSZOZUP7LCBOAVZ3XWJSXVANCNFSM6AAAAAA2VXJGI4 . You are receiving this because you were mentioned.Message ID: @.***>

[Dakes12]

Did you imported the CA keys into the harvester? …

Yes, I have imported ca keys as in chia tutorial.

[Kruger1981]

From your screenshots you are not farming as your farmer is syncing, until your farmer syncs and starts farming, your harvester won't work either as far as I know

[Kruger1981]

Also, is that log from the farmer or from the harvester?

[Dakes12]

Hello, Farmer log.log Harvester log.log

Find attached the logs from Harvester and Farmer. In this way it will be clear where is the problem.

Thank you again for your time!

[Kruger1981]

didn't have time yet to put together a new harvester and make the video, sorry, work keeps getting in the way lol

[NewSunSEO]

Hello everyone, I'm hoping someone can help me figure this Remote Compute out. I have three servers, one is the remote farmer with two remote harvesters. I want to load balance the farming between all three servers. How can I do this?

I entered in Linux:

./chia_recompute_server

This is what it shows me, but it doesn't let add the next part:

[chiapos] Using 12 / 12 CPU threads [chiapos] Using 1 / 1 CUDA devices Using 12 / 12 CPU threads Using NVIDIA GeForce RTX 3070 (CUDA (native)) ReComputeServer listening on 0.0.0.0 port 11989

I am not able to enter this to start the plots being sent between the servers:

export CHIAPOS_RECOMPUTE_HOST=10.1.1.1,10.1.1.2,10.1.1.3

Hey madMAx43v3r what do you think?

[NewSunSEO]

This is how I got it working. I have three Proxmox VM servers. On each VM, I started the recompute server like this in Linux FIRST:

If any of Chia functions were running, I stopped them: ./chia.bin stop all -d

I run this in Proxmox in a terminal window ON each VM: ./chia_recompute_server

Then, I go to that VM through SSH on my desktop & Run this command on each of the VM servers: export CHIAPOS_RECOMPUTE_HOST=10.1.1.1,10.1.1.2,10.1.1.3

Then I run either of these commands, depending on which machine I am starting Chia up on. I have my main remote farmer & two remote harvesters:

./chia.bin start farmer ./chia.bin start harvester