I’m sure there are few that do, but there always will be small percentage of cheaters. It’s impossible and unnecessary to accomplish 100% compliance. The key is that cheaters are in minority.
So no. Most SNOs are not doing this.
no, only who do not care anything but their income until they kill the source of the said income, like unhappy customers or lost segments because your setup suddenly disappear from the network on all 20 fake IPs, because your electricity provider decided that it’s the right time to fix your power lines for the next several days.
I don’t know. For my own storage node I am not upgrading my internet connection up front. In my case my ISP needs just a few days to upgrade me to a higher profile.
How about a TTL between 1-8 weeks? That would mean you need enough bandwidth to write between 25% and 200% of your free space per month. With my 100MBit/s connection I can write up to 32 TB per month. With a TTL of just 1 week that would mean 8 TB used space on disk and a 8 TB payout. With a TTL of 8 weeks I would cylce through 64 TB of disk space.
As I was going through these numbers my first reaction was “holy shit such a tiny payout for 100% bandwidth usage. That 1 week TTL must be a joke.” Later I realized that for additional 30€ per month I could go up to 1 GBit/s. That would be 320 TB to transfer and up to 80 TB payout even in the worst case with 1 week TTL. And there are some benefits to it. Instead of trying to grow my storage node over years I can get a decent payout in a short time.
Ok enough of the dreaming. I am sure these numbers are wrong in multiple places. Please do your own calculation for different TTLs and maybe don’t make the mistake to calculate with unlimited demand like I did now. There will be an announcement with way better numbers. I kind of feel bad to share my numbers with you so early on. Please wait for the official numbers before making any decisions. I don’t trust my own estimations enough to make any decisions myself.
How is this violating TOS? Please explain.
Maybe one more thought about the flaws in my calculations. Earlier I mentioned that we might end up with 75% of the nodes being full (!=75% of the total space). I am almost sure my node will be one of the remaining nodes with simply too much free space and not enough bandwidth. That kind of contradicts the second part about calculating my payout with a 1 week TTL. There are simply too many variables right now to make a clear prediction. All I want to transport is that as a storage node we will have to adopt to the new usage pattern. And even the worst case scenario I can think of still is a higher payout for my node compared to my current payout. So sign me up for it and we can figure out the rest later 
Edit: Outside of all the uncertainty there is one fact for sure. I will not upgrade my internet connection with the standard 12 month contract time. Instead I will buy the more expensive flex contract that lets me downgrade in just a month or so. I haven’t looked into these details yet.
oh, please do not repeat all this again.
at least this one
5. Restrictions. You will operate the Storage Node in strict accordance with the terms of this Agreement and in no other manner. Without limiting the generality of the foregoing, you will not:
…
- 5.1.7. Manipulate or alter the default behavior of the Storage Network to artificially increase or decrease the value of any reputation factor of any Storage Node;
…- 5.1.17. In any other way attempt to interfere, impede, alter, or otherwise interact in any manner not expressly authorized hereunder with the Storage Services or the operation of any other Storage Node(s).
…
You are not authorized to mimic several different physical locations by using a VPN to get more data in the one physical location, this directly affect the default network behavior (to do not store more than a one segment of the file in the same location). This is expressly forbidden. Please note - not authorized, it is forbidden.
This position never changed, even if I would prefer to have it implemented on the protocol level, than just in ToS.
There is always a bad alternative - by pricing level. More such cases and the time pressure would make the bad outcome is a simplest one, unfortunately. So in any case using VPN to bypass the /24 limit is the same as shooting yourself in the foot.
Using a vpn doesn’t ‘bypass’ the /24 limit. In fact is the only way for most SNOs to run multiple nodes in line with your rules.
According to http://storjnet.info the largest number of nodes are ‘hosting’. Since it is not profitable to use rented servers for storj, this can only be some kind of traffic forwarding. 
I believe this question is more about if it worth for the node operators to put more drives online, than if they are capable to do so. This is the same old question as ever…
For me personally, I have 10 available drive slots, but only 7 filled with active drives. These drives are 1x 2TB, 4x 4TB, 1x 5TB and 1x 10TB. I have 2x 5TB in spare, offline. These drives are filled with 16TB of data now, so about halfway, none of them are full at the moment. The peak was very close to 20TB with some of them full. This is with a residental 1000/300Mbit, single IP address internet connection. This gives me about avg 27-32 USD payout monthly. The house is running on solar, so I don’t mind the electricity cost of these…Yes the whole setup could be more energy efficient…
I used to swap the full node’s drives to bigger ones (move them or add additional nodes) whenever I can get my hands on cheap drives.
In my case the held back period is also a painpoint to add additional nodes…
The point is… I could easily bring more drives online but if I need another IP for every drive this doesn’t make sense.
Two out of my five nodes connect through VPN. Why?
Because the ISP where they are located has CGNAT and does not allow and/or wants money for port forwarding.
I use Oracle VPS to host wireguard bridge to connect those nodes. Yes, the geo ip is now different, perhaps there is maybe up to 200 miles difference between the node physical location and the datacenter.
Going forward most people will be behind NAT. Residential connections don’t need public IPs. This is the future. So I’d expect more and more nodes will be using some sort of bridges.
That’s why you have ToS. Not everything can nor should be enforced by technical means; the /24 ip rule is just a rule of thumb covering most common scenarios. Of course you can hack your way around it, but as Alexey said, why would you want to shoot yourself in the foot?
Are the new clients expected to be on the public or the commercial (select) network?
As far as extra capacity: At least on my end, capacity will keep growing as long as drives get filled up.
My feeling is the result would be the same if using just random node IDs. The /24 rule only makes things complicated for SNOs.
BTW
I guess the rule doesn’t apply for select SNOs? 
Total i have 512300GB is used by storj 328843GB I will find way to add more, when this will be filled.
I have several locations with 500/500 connection so speed will not be a bottleneck.
None right now, but I was going to buy two drives for personal storage anyway, just not yet—could as well do it earlier and store some Storj data. But I would also be willing to risk getting some non-SOC2 storage on the order of petabytes at a local datacenter if there was a substantiatied claim it would get filled reasonably quickly. I’d need about a week of lead time, maybe less. I understand it’s probably not going to happen explicitly though, just stating it as a possibility.
BTW, you can try recruiting some of the disappointed Chia folks, I see that a lot of them are giving up recently because apparently they don’t earn enough. Doing some legwork and marketing on places frequented by them could be helpful.
I suspect the percentage is smaller than that, but it would still be sufficient. The limiting factor would be ingress/egrees, not storage. Like, I can do 200 Mbps both ways and this is the top I can do with my current ISP while still getting enough bandwidth for my personal use.
Huge 
from me.
If the ingress increases abnormal high and new nodes are needed, we should discuss the collateral held by Storj.
The period should be reduced, along with the percentage held.
In 2023 I had 45$ held by the network for one node, and the ingress was high but not extreme.
If you get 150-200$ held by Storj for so many months, riots will start…
I think any modifications to the payout holding periods/percentages is quite tricky, but I believe it is worth looking into it to reduce the friction with creating new nodes.
The main issue (as far as I know) that results in having to retain some of the payouts at the beginning is to pay out nodes for repair egress whenever these nodes ungracefully leave or shutdown temporarily.
I still believe having some withholdings or similar system in place to incentive long term nodes is a good idea, but it could be tweaked.
The main way I could think to reduce withholdings (not eliminate completely) is to tweak the repair egress payout, such as:
In turn, then update the payout holding plan, such as to 6 months instead of 9.
While I’m not saying this is a perfect solution, and it would certainly displease some SNO’s, as payouts are a complex and sensitive topic, I think it could benefit the network as a whole by prioritising network health over the (at least for me) small payout from the repair egress.
With this said, there are other things to work on that I believe are most urgent/important than the payout holdings for new nodes, but it’s always good to keep the conversation rolling, and to keep it in mind if we start to look for new nodes if all of these customer deals go through.
Just me 2 cents.
Thanks.
However this one also leads to wrong used space display:
This node reports no free space but it has space available.
I like the ideea of collateral = repair cost. It’s predictible and fair. It could be hold like 100% of payout if the repair cost > payout; the exact ammount if the repair cost < payout.
If your node fills up and you allocate more, will start collecting collateral again.
The tricky part is when the node fills up and you decide you want to allocate less, because you overallocated. Should Storj hold that excess collateral? Or deal with a bounch of emails that request collateral back because they overallocated and they need the space, shrinking the node?
Hmm… maybe it can be automated to give back the collateral if the repair cost at the end of the month is lower.
I don’t agree with free repair egress. From my point of view, the egress is the same, I don’t know what it is - client data, repairs, audits, etc., so I want to be paid the same for each packet that leaves my storage. It’s already almost free at 2$ per TB.
Please tell me when to expect the start of full-scale testing?
In 1-2-3 weeks?
Agree with everything. After being told we don’t need enterprise grade drives, having them run for days on end at 100% usage is exactly what enterprise drives do. So are they still advocating using cheaper drivers or not I wonder….