I was mining Burst coin for quite a while. Now i have around 50ea of 4TB external HD.
What is possible solution to utilize all of this effectively.
Some intro
Best solution may be putting these HD in single machine like I used to do with Burst, but is that even possible? I believe one node is one HD and I don’t see anyone having 50 nodes in one machine.
On the other hand, putting HD across several machines will cost some more investments which I prefer not to invest further. I’ve seen a few mentions on Raspberry Pi which may be cheaper than Motherboard, but I have minimal experience with it.
Toolbox?
Do I need a toolbox as in https://github.com/TonyTosol/Storj-Node-Toolbox/releases
Many posts mention toolbox but I have not seen the full guide on how to set it up.
From my understanding, the set up is one common network, one machine with several HD - similar to what I want.
Any input on this and guide on setting this up would be much appreciated.
I don’t understand what it means? Does this mean you have 50 drives of 4TB each for a total of 200TB?! If so, it will never fill up entirely. Even 20TB only would take years.
Raspberry Pi4B are a good choice for power consumption. It comes with its share of drawbacks as for any system.
I’ve written something for helping me setting up a Node on RPi4B, it could help you, but it requires a bit of linux knowledge:
(Have a look at the README file - you may not want to use my utility storj.sh script - if you do, do it at your own risk!)
This might and should be true only for the time being. The outlook should be very different. If it takes years to fill laughable 20 TB then operating a node will not be worth it for many SNOs.
Storjlabs needs to move on rapidly to get customers on board to fill their cloud otherwise SNOs will not stay.
Good point, but it’s true if the number of SNOs stays stable.
If the number of SNOs keeps rising though, we’ll never go way above the current threshold.
But who knows…
@alphatio: It’s unlikely that you could possible have tens of different IPs, ideally distributed among different /24 subnets… But right now I believe that would be the only viable option for providing Storj with hundreds of TB.
For instance, you could have 4 nodes (4x4TB = 16TB) on 15 different IPs.
I agree. But data will increase exponentially. Today we have 4k, tomorrow who knows.
Youtube alone is said to have Exabytes of data already stored on their servers.
There is already a massive amount of data out there. Storjlabs needs to get those put onto Tardigrade.
The larger it gets the more market effects will come into play. If storage space fills up fast that will attract more SNOs too. And I for one will keep expanding as needed.
There is no reason to expect things will get better in the future and nobody should invest with that assumption.
Eventually healthy market effects will make it so that it’s worth it for SNOs with a small but healthy margin. To be honest I expect it’s more likely to get less profitable than more profitable.
It’s self-balanced system - supply and demand as well.
More SNO leaves - more data for remained - increasing profitability - attract more SNO - profitability start to drop and so on.
Also - more usage - more space will be available.
If prices are good and balanced - the system should survive.
It may not be encouraging, but look at it from this point of view.
Today if you buy an 8TB HDD you spend about $200. It takes about a year to earn pretty much exactly that much (more if you include held amount). It also happens to be filled up in about a year and from that point on it’ll earn you about $40 a month.
That’s a really good deal on a $200 investment. And unlike mining, you don’t have to worry about increases in difficulty or crypto currency fluctuations. Your investment is pretty solid long term.
As long as this is an attractive deal, it will attract more SNOs. But SNOs leaving will never be disastrous for Tardigrade. If free space on the network ever gets scarce, the above calculation becomes a LOT more interesting very fast for those who do have space available. A lot of existing SNOs will expand their setups and it would attract many more SNOs.
A good balanced market automatically finds an equilibrium where everyone is paid what their service is worth. That will happen here as well. I think it’s good to realize that we’re in a luxury position by having joined early and we’re doing pretty good at the moment.
I agree. Long-term I’m a bit afraid it might end up a bit like mining where only the countries with low electricity cost are still profitable but I hope that doesn’t happen to storj as with HDDs the electricity costs have a lower impact.
@kevink I use less than 15W on Raspberry PI 4B (8GB) + 2 TB HDD, so I think it will be even possible to power it using sunlight with some batteries. Raspberry PI is so cheap (and one time investment), but it is able to function using very little power. Before I had to use my stationary computer that was using up to 450W (Respberry PI has 30 times less power consumption, and this is including HDD).
Agree with your post but doubt this number unless your PC is ancient. The PSU power is not the power draw. A modern desktop only uses 450W when you are playing games but not when only idle with storj. my desktop for storj uses 60W with 3HDDs and my desktop for working uses 60W when idle with 1hdd.
You can run as much nodes as you want. But all of them will receive in summary as a one node - we select one random node from the /24 subnet for each piece of each segment of the file.
