Hi Captain_plots-a-lot ,
Thanks for your answer . I will try these and come back .
If anyone else can direct me to a ( nearby ) solution , that would be kind .
That’s what made me think it’s hardware related, but perhaps it’s not.
1 Like
If you go into disk manager and can’t see the disks, yes you have a connection issue. Are these connected via USB and they are asleep???
Hi guys,
I’ve been re-plotting slowly using Max CPU plotter.
I’m farming with a PC using flexfarmer on WSL with difficulty 20 on a windows 10 machine and using USB drives :
a 460 TB farm that now looks 539 TB thanks to C8 plots.
The GPU farming is a GTX 1060 6GB (undervolted, slighly underclocked).
The CPU is a E5-1620 v3 @ 3.50GHz with 64 GB RAM
My issue is that I tend to see lots of duplicate partials these days, like that :
And I don’t know what to think about it.
On the pool side, I’m quite a good farmer, usually with 0 stale, 0 rejected :
I would like to know whether it’s going to be worst or not with c8 plot population growing, and what action I should take if needed…
(@Chris22 and @madMAx43v3r you’re requested here!)
1 Like
Duplicate partials can only be from two reasons:
- Running more than one farmer, ie. two flexfarmer or flexfarmer + chia farmer
- Accidentally copying the same plot multiple times to different destinations
1 Like
Also chain reorgs cause that error I think but since there is no reorg…
1 Like
I definitely do not have any accidentally duplicate plot.
I’m definitely not running more than two farmers for what I know…
But, is it possible WSL does? (it’s been running on this windows session for a looooong time)?
Just to be sure, I shutdown WSL completely :
wsl --shutdown (as administrator)
and restarted Ubuntu/WSL and flexfarmer…
And guess what? no more duplicate partials! 
BONUS : what difficulty would you suggest for such a farm on flexpool? (460TB of c8 plots)
2 Likes
Depends how much compute margin you have, between 18 (for better stats) and 100 (to save compute).
1 Like
Hi drhicom ,
When gigahorse farmer is not turned on , I can see EVERY HDD . I can also write/read without any problem . I let the the computer turned on for 3 Days and I have no issues .
It is only when gigahorse farmer is online that I have that problem ( cant detect them in disk manager ) . They are all connected with SATA port .
What versions of the software are you running?
Are disks mainly sitting idle during this time?
How are they powered?
I don’t know if this will be helpful or not but this is what happened to me. A while back even before compression I was growing my farm and kept adding drives to my farm when my farm started doing the same thing. My problem was hardware. The problem I had was that I had passed my 5v amp rating of my power supplies. So on spinup and idle my farm looked good all drives detected and individually worked as they should have. But when the chia client would read multiple drives simultaneously looking for proofs, drives would drop. And would refuse to be recognized until a reboot of windows. My fix was to add more power supplies to keep stable 5v power for times when disk access was happening. All ATX power supplies have a far lesser 5v rating than they have 12v. So for instance If you have an 800 watt ps and only 36 hard drives that consume 12 watts each max then you should be good, right? Wrong. Even if 36x12 is only 432 watts you must remember that not only does your system have other overhead but if you look at the ratings of your PS and check the 5v rating it will probably say maybe 24a. With few exceptions every hard drive has a 12v amp rating and a 5v amp rating.
The 5v side for almost all hard drives is used for read/write. The 12v side is used to spinup and keep the drive spinning. So if I look at my 5v amp rating of my 32 hard drives and each one requires for example 880ma. 32x.880 = 28.16a which exceeds my 5v rating on the power supply which has in this example only 24a on the 5v rail.
Sorry for being so verbose……
9 Likes
while low-wattage power supply unit sound pretty reasonable let me add my two cents
that is also might signals this is something wrong with south bridge on your motherboard. just touch it with your finger, pretty good test hehe …
moveover, i would say PSU issues leads to overheating of SB and such killing it in a bad way.
Hello Max !
It will be soon 1.8.2 version of farmer or 1.8.1 will be the latest for while?
2 Likes
4 Likes
Cheers Max, good to see an update and hope you are having a nice summer.
Just FYI this time everything is working flawless, last time i tried it was 1.8.0 and i had trouble running remote harvester on Windows Server 2019. I’m not sure was it me chia itself or something else but it does work now. Thanks!
1 Like
Hey @madMAx43v3r and all.
Let me ask this, how come farming of just 10 compressed plots eats up to 6Gb of VRAM?
See that K33 C8 you have there? …that’s why. Doesn’t matter whether you have 1 or 10,001 of them, that’s what it takes, approximately.
2 Likes
oh wow i see. thank you. totally forgot about this table.
and yet i wonder why memory isnt freeing up eventually when expansive k33/c8 check is over for current plot challange. it keeps being occupied forever.
1 Like