I am sorry to have to disagree with you, @Alexey
IANAL but unless the official ToS have that very statement then I’m not sure your position could be enforceable in a court of law (although this is a bit of a “ad absurdum” argument, no-one is going to court over running multiple nodes under the same IP)
Also, from a practical point of view, you cannot expect people to have to search through this forum to find out which points of the ToS have been superseded or amended.
Ideally, the official ToS should have been changed (about 4 years ago according to this post)
Before lowering the rates, Labs conducted interviews with some SNO, and in response to my question about the /24 rule, they mentioned that there have been internal discussions about this rule within the company, too.
That’s why I was interested in knowing how strictly they enforce this rule currently.
The /24 rule stays. I have never heard any noises to suggest it was even being considered to being dropped.
The one-node-per-IP is mostly being ignored and eventually (some time between now and the heat death of the universe) will disappear.
I think that pretty much sums it up
I appreciate that you removed your comment about what would happen in court. Because courts not only look at what you said, but what you did.
IPs used by the largest VPN providers are publicly available. Storj is aware of them, and they have not taken action to block them. Instead they’ve configured their satellites to hand out Node IPs on those VPNs to clients, then charged them for storage and egress from those VPN nodes.
So while some text-on-a-website and a rep-in-a-forum may say VPNs are prohibited: the business is actively and intently making money from them. I agree with @ACarneiro that such case would be absurd to bring in front of a judge: because Storj would get dismantled. They’d be laughed out of court.
VPNs are allowed by the terms of service.
Running multiple nodes behind a single IP is allowed as well.
What @Alexey is getting at is running a node on a native ISP connection + routing another node through a VPN in the same location. This increases the risk (depending on how many nodes you have ofc) to store the same piece on all of those nodes. This can lead to a piece disappearing (=lost) if your hardware fails.
Personally I don’t view running a native ISP IP + routing one through a VPN as a violation of the ToS as long as both(or more) are each on separate hardware, running on a separate power source (+ having a generator). Technically speaking that’s no different to owning an appartment building with each appartment having its own network connection. Yes there is the chance of the building coming down due to an earthquake, but it most likely wont.
If you don’t have alternate power then if the power company decides to take down the power grid in that area for maintenance/repairs it could mean lost segments.
The ToS could be tweaked here, IMHO: each SNO is allowed to run X nodes in the same location, and only if those nodes are running on separate hardware. This rule shouldn’t apply for nodes running behind the same IP, since the satellites already counts them as 1 node so they can never store the same piece at the same time.
I don’t know, that’s why it was an “X”. Storj can pick a value for X according to their required durability, ie taking into consideration how likely it is that all nodes running in the same location would disappear.
What I see happening is latency becomes very high, and some repairs fail. Audit score drops as a result of failed repairs.
That’s not the only reason to use VPN. My two nodes connect via VPN because ISP does not offer public IP.
Is it? Can you please point to specific language?
Interesting. How is SNO expected to read ENTIRE forum to maybe find amendment to Tos?!
Anyway, do you want me to shut down my two VPN-only nodes?
This I agree with. It would be dumb to shoot oneself into a foot by degrading network reliability, breaking assumptions, let alone ToS. But VPN alone should not be a problem.
Again, if it is – please add this to ToS and I"ll delete those two my nodes.
This is well-known datacenter with very reliable equipment. They do not need to use a VPN to bypass anything, they simple owns separate networks.
The problem with bypassing the /24 rule is in the fact, that several pieces of the same segment may be uploaded to nodes, which mimic different locations. If that single location would have issues (with electricity or internet or hardware) these pieces would become unavailable. This increases risk of unrepairable segment, this would mean that data is lost even temporary (while such SNO fixes issues) or permanently (the disk is died).
The lost segment mean the lost file. The lost file = lost Customer = no payment = no payouts.
If you do not have a datacenter-grade setup with a redundancy, you should not bypass the /24 rule.
If you have a public IP, you should not use VPN.
Regarding this topic - the author has a public IP but have also several VPN to bypass /24 rule and these VPN connections are have issues with uptime. So, the obvious solution is not use VPN.
You are right here. We still need to update ToS to have everything in the same place.
The new version has been prepared a while ago as far as I know, but didn’t pass audit so far. I would again ask the team about the progress.
