at least if plotting wouldn’t be so expensive and taking so much time…
there are folks selling plots starting at $1 each, apparently that does not seem to be the scarce resource for chia. The growth rate of space is enormous, more than 1 million 10TB disks already…
Was just looking at goofys which has no data cache, wondering if it would work…
Only one way to find out. Feel free to give it a try. I was looking for a windows solution myself though. I’m going to consider my experiment done. I’ll keep it running for a while longer to monitor if it keeps working, but I’m done tinkering with it for now.
What does the logfile say about the time it takes to read those 2 plots from storj?
I do have a few warnings in the logs that looking up qualities for these plots is taking too long. It should be below 5 seconds, but tends to take over 20. So I guess there is a pretty big risk of losing block rewards as a result of that.
Hmm yeah that doesn’t sound too good. So mining from storj might not be fast enough but without knowing how often it happens, it’s difficult to say if it could work or not.
Locally I set up a monitoring stack with prometheus, grafana, promtail, mtail and loki. I can see that looking up qualities only takes more than 5 seconds if my HDDs are really busy with heavy copy operations. But setting that monitoring up just for the storj experiment is probably a waste of time 
Considering how slows S3 Drive is with other things, I’m not even sure Storj is the bottleneck here. But I don’t feel like looking for yet another way to mount S3. To do a real test we really need a native way to mount Storj to local storage.
I do also see these warnings sometimes for the plots I have on my NAS over SMB. But the times are just over 5 seconds in those cases, with a max at around 10 seconds. Still, Storj over SMB > S3 Drive > S3 gateway MT isn’t even that much slower it seems.
I had the same idea but the problem is mounting in windows, I had used rclone but its very unstable and pretty much crashes the entire system when upload the 106gig plot… Also the program wouldn’t even let me use that mounted drive.
Not sure I understand your obsession with windows. And Mr. BS’s.
Windows works well where it works well but that doesn’t really include non ntfs/fat filesystems.
Theres no obsession with windows Its what I use on a daily basis everything just works, Linux theres always configuring and half of the time it doesnt work.
Sorry, I’ve wrong worded again. I mean to say that I … meh whatever Windows for the win!
1 Like
No worries also ive been unsuccessful in uploading 106gig file to storj so far. let alone running it off chia…
1 Like
It’s just what I had on hand. My only Linux system at the moment is my NAS, which doesn’t really like it if you install stuff outside of the Synology store. Trust me, for this stuff I much prefer Linux. My first attempt was to use WSL for a reason, but this was supposed to be a quick experiment and I didn’t exactly feel like spinning up a different machine or VM for it. I got out of it what I wanted though, so I’m happy.
1 Like
Could I pay you some storj to persue it further?
Chia is up a above 10,000 petabytes and even a small cut of that would fill storj dcs.
I’d offer hosting but my broadband can’t
It is a bit tragic that all of this space is not used for anything else than farming Chia. No real world usage behind it. Just imagine if Storj would collaborate with Chia and somehow use that space to actually store something while farmers would still earn (whatever it would be Chia or Storj or both).
That is an interesting idea and brings me to the question if Storj is actually used anywhere to pay with besides here?
As Storj|DCS aims at developers now, it feels like it could be a pole position to pursue adoption as developers could not only implement Storj as storage solution but also as form of payment. Maybe even websites for freelancers etc. could be attained to start accepting it.
3 Likes
To what end? When I said it would be profitable in theory for at least another week, that was only part joke. It won’t take much longer before it is no longer profitable. And I don’t have the SSD’s to do the plotting on anyway. As a proof of concept it was a nice experiment, but it’s not viable for a normal Storj DCS user. I was suggesting Storj Labs might want to use it as a test case and to keep SNOs happy, because they could have some reasons other than profit to give it a try.
If you’re interested in this use case feel free to give it a try yourself. I would be curious what the response from Storj would be if you ask for a limit increase for this use case though. 
1 Like
I wonder whether it would be a good idea for Storj to use Chia plots instead of existing test data…
the main problem i would see with this, is that on the Storj Network data is expanded by 2.7 or whatever, so essentially any data stored would cost 2.7x the price of storing the data on hdd’s
so tho it’s possible it will never make sense or payoff long term… except if you like created a petabyte or whatever in plots until you get into the lead and the gets a lot of blocks and then is lucky to be able to sell XCH at the crazy prices it’s at
and then you either get out while the going is good or you spend the earnings on hardware to sustain the behemoth of plots on a local system.
ofc XCH prices might drop… they just need to half before it basically earns about what it costs… so thats already kinda tight for a long term project because you need to generate a MASSIVE amount of plots…
it is a crazy and most likely fail scheme… that will never practically make sense for anything other than a temp fix… which still might fail
windows mounted storj cloud is pretty cool tho
If you look at their reference specifications, detailed out here
and here
I tend to agree with you. You don’t only need high-end machines for plotting, you also need tons of hard drives for farming. How does this make sense?
One of their goals was that everyone could farm on a laptop. This has turned into crap as it seems as with the current network size it requires years of single farming - eventually longer than a decade - to find a block.
How is all that ‘everyone’ and how is that eco friendly (maybe compared to Bitcoin mining but is it really ‘green’)? And I don’t even speak of the tons of SSD that get turned into bricks because of massive plotting. (BTW buying used SSDs does no longer seems to be a good idea, if they had been used for plotting before huh?)
Yes, maybe Storj will get someone to code something that is stable with great performance or improve some existing solution.
3 Likes