How do you save your plots?

hi guys I am just wondering how each of you is saving their plots? please state your method let us learn from each other.
I started with HDD USB Bays but they quickly filled up. now I am using JBODS with NAS server and it is a little hassle for me to transfer files over the network to make space for the new plots as it is slow and can ot keep up with the plotting speeds. but this is my way.

thank you

1 Like

Can your nas run the Harvester code on it?
Many ppl had issues with nas and poor speed performance.

I use external USB caddys, and internal hdd.

I have both NAS and JBOD. I don’t have any issues harvesting over the network. You just have to have a harvester for each device (even if it is remote).

To answer your question, I have 3 plotters. They save the final plot locally and then I have a Microsoft Power Automate script (flow) that goes through all of my storage drives (current 40 different locations) and finds space for it. Then moves the plot to its final location.

1 Like

I started with External drives as well
4,6,10,12 etc…
after 10 of those… it seemed. Messy?
the next 5 drives are all Seagate 16tb internals.

I have 4 machines plotting as needed
┌i9 9900
├i7 9700
├i5 10600
└R9 5900X (hosts drives)
I am going to look up how to split farm/harvest machines “soonish” to get some better performance as the network grows!

130 8tb USB drives…oh the wires…

Glad it works for you, was reading this thread, and put me of looking at nas.

I participated in that thread a lot and it actually branched off to a new thread that talks specifically about using a NAS as storage. I did a bunch of testing and found that if you have your harvesters setup properly, there is no problem using a nas. I have over 2500 plots on nas devices and no problems.

1 Like

I see one post by you in that thread, not my definition of alot.

Perhaps you were more in the branched thread.

I plot on internal hdds some shucked from externals.
And just relocate the drives a few at a time when they’re full. Cant be bothered with all the copying.
I also have a few hotswap bays in each machine, that help with that

1 Like

I posted a lot, then they took my posts and broke them out into a new thread because they were more focused on a single topic (using a NAS).

2 Likes

When I was actively plotting, I plotted to an internal HDD. Every once a week, I would go in and transfer plots to my NAS/farming server over the network. I tried transferring plots to a USB drive then to my farmer but that was almost the same speed. I tried plotting to my NAS directly and that caused issues with the speed of plotting. What I did not try was a robocopy script or a Automation script, but I assume based upon regular network speeds that it would be the same. The only benefit of robocopy or an Automation script would be automated moving of the plots.

what is the speed you have over the lan? i have connected my network on dual port 1GBIT network cards still it run a little slow around 100Mbs/sec
can you share your network setup?

probably yes but i am not a Linux guys so i am not sure

Depends on setup, i know ur post wasn’t aimed at me, but 5 bay usb docks, that max out at 16tb per disk ( even 18 tb are truly only 16) , is 2 cables, and 2 power leads, not so bad.

My network is all 1Gbps. But I have it split up between farming and plotting traffic.


This image is a little out of date, but the network structure is still the same.

1 Like

Background: esxi with 4 Ubuntu 20.04 VMs. 1Gbps network.

On the harvester:


On each of the 3 plotters:

On Swan config of each of the 3 plotters:
image
Once a drive is full simply delete the correspondent lines from Swan config.

Average time check is good (0.7s). Occasionally you’ll get spikes when the harvester reads while a plotter writes, but the overall stats seem acceptable:

image


6 more internal sata drives in pc case.
I plot on my main rig via usb (external enclosure if it’s an internal drive) then move full drives to this harvester pc. It works for now anyway.

2 Likes

Love the lego stand lol