search

madMAx43v3r / chia-gigahorse

220 stars 32 forks source link

6,000+ C18 GIGAHORSE PLOTS - SPACEPOOL ONLY SEEING ~135TiB #246

Closed NewSunSEO closed 7 months ago

NewSunSEO commented 7 months ago

Hello, I am in the process of re-plotting from C8 to C18 Gigahorse plots. The C8's, SpacePool was seeing all the plots no problem (Was up to ~700TiB before re-plotting). I got more HDD's and started re-plotted to 6,350 C18's so far, but SpacePool hasn't shown more than ~150TiB of plots since I started re-plotting a week ago. The plotter has been creating new plots 24/7. Right now, SpacePool is only showing ~134TiB of plots, which it hit ~150TiB the past two days, but has dropped back down to around ~135TiB all day today.

This is my current setup, I'm guessing something needs to be upgraded for these C18's like maybe the GPU I'm farming with, but I am not sure. Maybe the 3060Ti's can't farm these C18's? I have another 3070 I'm going to try in this system tonight. I'm hoping this gets updated soon: for the GH stats https://xch.farm/max-farm-size.

PLOTTER-01 / FARMER-01 Lenovo Thinkstation P620 Threadripper PRO 5945WX - Proxmox OS SeaGate FireCuda 530 SSD's RAID 1 for OS 512GB 2666 RAM P4000 (Plotting GPU) 3060Ti (Farming GPU) FARMER-01 is setup in a Proxmox VM running Ubuntu Desktop / 32GiB RAM - (~3,808 Plots Currently)

HARVESTER-01 Dell R720XD 24 Core 2X E5-2697 V2 - Proxmox OS Samsung 870 2TB EVO SSD's RAID 1 for OS 512GB RAM TrueNAS - 12 20TB SATA HDD's RAID 0 - In two pools, 6 Disks each pool 109TiB each pool / 218TiB server total 3070 (Farming GPU) HARVESTER-01 is setup in a Proxmox VM running Ubuntu Desktop / 32GiB RAM - (~1,271 Plots Currently)

HARVESTER-02 Dell R720XD 24 Core 2X E5-2697 V2 - Proxmox OS Samsung 870 2TB EVO SSD's RAID 1 for OS 512GB RAM TrueNAS - 12 20TB SATA HDD's RAID 0 - In two pools, 6 Disks each pool 109TiB each pool / 218TiB server total 3070 (Farming GPU) HARVESTER-02 is setup in a Proxmox VM running Ubuntu Desktop / 32GiB RAM - (~1,271 Plots Currently)

NewSunSEO commented 7 months ago

I'm beginning to think these C18 plots are too much for the 3060ti, but I have not found anything about this, other than this link that does not have GH stats in it yet: https://xch.farm/max-farm-size

It sucks to have well over $15k invested in this current setup, which is barely bringing in $1 / day currently. It was doing almost $5 / day with the C8 plots.... Which even then would take 5+ years to ROI on at current levels....

NewSunSEO commented 7 months ago

I have the 3070 in the system now. It is rebooting. I will update everyone with my results. I am not going to plot more C18's though, until I resolve the current issue, or I go lower in compression. I have a 3090, I just can;'t fit it into this current configuration...

NewSunSEO commented 7 months ago

The 3070 has been in the farmer for over 13 hours now. SpacePool is only seeing 120TiB... It should be seeing over 400TiB (Effective) though... Is anyone else having this same issue with SpacePool only seeing a fraction of the C18 plots that it should be?

Kruger1981 commented 7 months ago

I am using a 2080Super to farm 5000 c18 and another 6000 mix of c8 c7 c5 and c0 no issues, your problem may be you are using your farmer as plotter and getting a bottleneck somewhere. Test using ot just as farmer for a day or 2 and see if plot space goes up in space farmer

NewSunSEO commented 7 months ago

I took the 3060Ti out & put a 3070 in. Upon more inspection, I did not have the GPU passthrough correct. I was passing through the previous GPU that was in the system & not the 3060Ti. I put the 3070 in anyway & corrected the passthrough & the plots started climbing again. It was farming with the CPU I am guessing, but I got it figured out now. Thank you for your comment though Kruger1981.