@echo off
:RESTART
TITLE Move Plots from Drive1 to Drive2
echo.
ROBOCOPY “E:\mmxplots” “P:\mmxk32c” “plot*.plot” /MOV /R:999 /W:5
echo.
timeout 10 /nobreak
GOTO RESTART
pause
@echo off
:RESTART
TITLE Move Plots from Drive1 to Drive2
echo.
ROBOCOPY “E:\mmxplots” “P:\mmxk32c” “plot*.plot” /MOV /R:999 /W:5
echo.
timeout 10 /nobreak
GOTO RESTART
pause
I propose that the following line be changed:
From:
ROBOCOPY “E:\mmxplots” “P:\mmxk32c” “plot*.plot” /MOV /R:999 /W:5
To:
if exist “plot*.plot” ROBOCOPY “E:\mmxplots” “P:\mmxk32c” “plot*.plot” /MOV /R:999 /W:5
The above will keep robocopy from running when no plot file exists (reducing repeated screens of output from robocopy).
hello dr hicom
Thanks but no work in my windows iam noobie XD
I create .bat and copie your text
D:\Plot is my Source and G:\ is my HDD
@echo off
:RESTART
TITLE Move Plots from Drive1 to Drive2
echo.
ROBOCOPY “D:\Plot” “G:\” “plot*.plot” /MOV /R:999 /W:5
echo.
timeout 10 /nobreak
GOTO RESTART
pause
I made a mistake ?
It looks correct.
Please copy/paste the output here, for us to evaluate.
Not really, I did think about making it public, but then if something goes wrong, keys are stolen etc I could get the blame, and I don’t want that, sorry.
You have to edit your drive letters and paths.
That Working now thanks all
Hi, this is a silly question…I did the wget per above and when I try to do nano plow.py, everything falls on one line? Seems like I’m missing page breaks or something. Also copied the file to Windows (where I am more comfortable), but same thing. What am I doing wrong? Thanks!
I wrote my own plot mover for Windows which has a GUI, which is available on GitHub.
Thanks for the help Ronski bro.
Transfer plot from DELL 720 linux machine to PC harvester rig. I made some progress. I obtained a SFP+ card for the Windows PC, an SFP+ cable. Direct connect a linux to windows 10. Did some simple configure learned for youtube. Now I am doing SMB protocol sending data from linux to windows. Two files simultaneously. 100MB/sec each, sustainable. Way better than gigabit home network.
I am sure I am far from the optimized situation. This is a good step forward.
Two Dell 720 cuda plotter are doing C5 Bladebit plot now. Cheap 1080 GPU. 5 minute a plot each machine is doing.
Thank you very much. I finally figured out PLOW. Finally my distribution speed can match the plotting speed now.
Update to this old post.
I have a R720 with 12 3.5 bays with 1080 GPU. I made a 12 SATA RAID 0 (12x4TB) for plotting temp and dump. It was not as fast as I expected, espcially sending out files to desitnation hard drives.
Last weekend I changed these 12 SATA HDD to 12 SAS HDD. Same old and cheap Seagate ES3 model 4TB, just SATA vs SAS.
SAS flies. At least 2.5x faster over all.
for our part, we are creating a wrapping around bladebit cuda.
we are kind of pissed of about bladebit missing some vital feature like
As you can see, this is not yet merged (for the stagging part, for the other it’s on master).
We are running it under windows & linux, 128/256G. binary provides was for us (we patched logs output for windows to [add timestamp](https://github.com/Chia-Network/bladebit/pull/376)
to it), but you can use your own if you want
if it can help.
multiple destinations are possible since alpha3 if I recall bladebit_cuda ... cudaplot dst1 dst2 dst3 dst4
at least in Linux.
what you mean by staging, please?
no really ? well god news
stagging is like plot super fast with sdd, then another thread will copy on output disk.
all bottleneck come from disk write speed, like this if you have more than 6 hdd to plot, you can plot 6 plot fast on ssd, copy then, and keep plotting directly on disk.
Once one plot get out from stagging, you will plot into again.
We are at 3-4 minutes to plots on sdd, and 13-20 to HDD (including copy over network for both).
EDIT: does it keep crashing if it cannot write the whole plot or go to the next directories ?
I used always 4-8x HDD stripped JBOD - it does 600-1000MBs minimum.
So far, I didn’t run out of the space…BB used to crash if no space left at destination
The latest versions of bladebit no longer crash, they keep checking every 10 seconds to see if space is available. I don’t know what happens if you have multiple destinations though.