Update Proposal for Storage Node Operators

Announcements
1

Hello Storj Community,

We have some proposals in the works for changing the payout structure. We’re looking for feedback on these proposals, if it doesn’t work for you, please talk to us.

This is a proposal, not an announcement. There will be no abrupt changes, or surprise changes that we have not communicated and requested feedback on well in advance.

The network isn’t possible without people like you that provide reliable capacity for the network and we appreciate the time, effort, and resources you have committed to the project.

Overall, we need to make sure that participating in the network continues to be economically viable for both node operators and satellite operators.

What changes are expected in 2023?

Reduced Synthetic Load

We plan to continue decreasing the synthetic load on the network at a pace that matches the growth in customer data. So each month we will evaluate the amount of customer growth and decrease the synthetic load to match. If the growth slows, then the decrease in the synthetic load will also slow, and vice versa.

Evaluation of the Future of the Free Account

During the year we will also consider a restructuring of the free account. We are not yet sure where this evaluation will lead, but it could mean reworking the free account and/or lowering the free tier, transitioning to a free trial, or something else. We are feeling like currently the free tier is not being used for the goal for which it was intended. As such we will likely make some changes to the way it is structured.

Adjust Payout Unit Economics

The goal of the Storj project is to create a free market where nodes and satellites engage and find an equilibrium price for the services that the SNOs provide to the network. The current state of disequilibrium not only has the consequence that satellite operators (Storj Labs today but others in the future) are overpaying for capacity, but it also means that the network is over supplied vs the demand that we currently have. This oversupply actually negatively impacts the economics for Storj nodes, as it decreases the rate at which nodes receive more data thus increasing the time it takes for nodes to reach max capacity, and thus reduces the lifetime payout that is possible for a node to achieve. In order to address this oversupply we are proposing reducing the amounts paid to nodes for storage and egress.

As we adjust down the payments that we provide to nodes we understand that this will mean that operating a Storj node may no longer be economically viable for some SNOs. While we will of course be saddened if these SNOs choose to stop contributing to the project, we know that adjusting these payments is required for the long-term health of the project, and hope that it will drive creative and innovative solutions for the operation of nodes in an economically efficient manner.

In proposing new payouts for SNOs we do not presume to say that this is the “equilibrium” price for this market. We are simply offering a data point based on the understanding we have of our business as well as the feedback we have received from many SNOs. We fully anticipate changes in the future that may necessitate us to later increase or decrease these payouts as the market adjusts.

We are proposing the following ranges of potential compensation amounts to nodes. We hope that at these levels the economics for nodes are still appealing, but want to get your feedback to make sure we strike the right balance before a final payout is determined.

Current Payout Proposed Payout
Storage (per TB Mo) $1.50 $0.75 - $1.00
Egress (per TB) $20 $1.50 - $5.00
Audit / Repair traffic (per TB) $10 $1.50 - $5.00

While the proposed reduction in the payout for bandwidth egress may look disproportionate to the amount charged for egress, there are a number of factors involved. The majority of egress bandwidth is served through edge services (Linkshare and S3 gateway). These services did not exist at scale when the previous payouts were set, and the infrastructure for those components includes significant costs that we needed to incorporate into the model. We’ve learned a lot about the costs to scale those services since they launched which is now being factored into these proposed payouts.

When we do change SNO payouts, whatever the future amount is, we are planning to roll it out in a gradual way which will include some sort of surge payout while the network growth catches up. We want to avoid any sudden drop in payouts. It will also give us time to evaluate how well it works for the community and respond to ongoing feedback. Given that our initial plan is to transition to the new pricing with surge payouts within the next month or two, then the level of surge pricing will gradually reduce as the demand grows.

We look forward to hearing your thoughts and feedback on these proposals. Please add your feedback to this thread, or join us on upcoming Twitter Spaces.

Thank you,

John

8 Likes
2

Sorry for the dumb question. What is synthetic load?

2 Likes
3

Not a dumb question - synthetic load is our test data that simulates customer data. We added a bunch of test data and egress usage to bootstrap the network. Now that we’ve got a steady increase in paid customer data, we’re reducing the test data/synthetic load.

7 Likes
4

I knew that the amount would be bad but I never expected that the prices would be this bad, I get it though to make money you gotta spend money but at some point you gotta stop over spending to make money again.
Im pretty sure this will stop many SNOs in there tracks to where they would just stop running there nodes not even caring about running GE. Some SNOs have super high elec costs to where this wouldnt even cover much of anything especially exiting the network.

Current Payout Proposed Payout
Storage (per TB Mo) $1.50 $0.75 - $1.00
Egress (per TB) $20 $8.00 - $10.00
Audit / Repair traffic (per TB) $10 $1.50 - $5.00

I think this amount would be a more realistic if you wanna keep some SNOs running on the network. But its still a loss for some SNOs.

7 Likes
5

this is almost an insult…

sure storage prices will keep dropping and i guess these values would stay for a good number of years. most likely… without any major technological disruptions.

but these new proposed numbers would be something like
60% to 90% reduction in earnings for SNOs

i don’t see how that is realistic, maybe in countries where power and man hours are much cheaper…

i was expecting to see something like.

Proposed Payout
Storage (per TB Mo) $2.50 (tho i suppose 2$ could also be acceptable, since storage prices will without a doubt keep dropping as time passes)

Egress (per TB) $5.00

Audit / Repair traffic (per TB) $1.50 - $5.00 (this is from what i can see just like 1/5th of my egress so not going to change the math much)

i think raising the TBm payouts is the only way to make up for the big losses from the massive reduction in egress earnings, this would also put more of the pressure on storj to work towards getting more active data on the network…

holding backup’s for people are fair enough, but really not where the best money is, anyways…

13 Likes
6

Who would have thought that the two main characteristics of Storj ( E2E encryption & Decentralization ) are NOT something that the market in general want/value enough.

Maybe we SNOs need to “pack our things and go home” and you can keep providing the service a la AWS et al.

15 Likes
7

I understand that something needed to change to keep Storj viable. That’s why I started my previous topic to begin with. I even understand that my guess at what payout could be in the future may have been a little optimistic. I mentioned as much in that topic. But I am a little disheartened by the large range of future payouts proposed below that $1/$5/$5 suggestion. Especially on the egress side of things. I was really hoping these payout changes would come alongside cost saving changes on the network, like tuning RS settings to lower the expansion factor and better incentive structures to limit repair. These kinds of things would make the network more profitable for everyone involved as there is less wasted on overhead that nobody benefits from.

This part I understand. Running these services is expensive as these middleman services incur egress charges on both customer uploads and downloads. But at the moment there is little that incentivises customers to use the much less wasteful native integrations. Furthermore, these services eliminate some of the unique advantages of Storj, like practically unlimited scaling through parallelism, decentralization of traffic flows and client side encryption. They effectively reintroduce the traditional cloud storage bottlenecks.
It would be beneficial to give customers an incentive to go native by using price advantages. For example, dropping egress costs for native connections to $5, while adding a $3 cost on top of that for upload and download using these edge services. There would be no reason to pay node operators any less than $4 for egress in that case.

So what would this mean for my nodes in practical terms? As outlined in this post assuming the upper range of suggested payouts listed here, a purchase of an 8TB HDD barely makes sense any more with the current energy prices. And even with a 20TB you would effectively only double your investment in 5 years time.
Ok, well, Storj has always said we shouldn’t buy new hardware, so what about when you already have the HDD. Well in that case 4TB would be barely profitable at these payouts with my current energy costs.
That said, the cost for energy is about to drop for me, so 4TB is probably still going to be worth it for me. However, I run 8 external HDDs with nodes smaller than that. And those will likely run at a loss with even the highest proposed payouts here. This means I have have to consider shutting those down and sticking with only the nodes internal to my NAS which was always on to begin with.

Keep in mind, I’m ONLY accounting for the energy costs of running the HDDs themselves as in my setup everything runs from that same always on NAS. Other people who run separate servers for their nodes or even just raspberry Pi’s would have higher costs.

Partly this is just because of unfortunate timing during an energy crisis. But it does show that these high limits of this proposal already scratch up against the limit for profitability. And even though my setup is really energy efficient compared to some others, I’m gonna have some choices to make.

  • can I still afford to buy expansion when my disks fill up?
  • can I afford keeping smaller HDDs running for Storj?
  • should I find a cheaper power company (yes, I should, I’m working on this one, haha)

All in all, this is partially what I was expecting. But the lower limits of this proposal seem very low to me and I was really hoping for a more holistic approach that went beyond just lowering payouts. Given that Storj still has opportunities to optimize the network to lower costs, I would argue that squeezing profitability out of just lowering payouts right now is probably not entirely fair. So I would suggest sticking to that upper range of this proposal and use the limited profitability for Storj Labs as an incentive to further optimize the network to make it less wasteful.

25 Likes
8

Storj Labs would lose $1-3 per TB egress with this suggestion. And that’s without accounting for cost of edge services.

Again, due to the expansion factor this suggestion (even the $2 one) would lead to Storj losing money per TB stored. Further more, this is more than we get now. I’m not sure what you were expecting, but higher payouts were never in the cards for this change.

3 Likes
9

Well… we were expecting to get bad news, but for me and I think for a lot of SNOs the most important is storage price. Drop in price for egress was expected, to let’s say 5$ but storage should really stay at 1.5$. Storage Price is equal to steady income, and if that goes down by 50% then rip, because people won’t have enough earnings to even pay for the electricity.
Those proposals will rip a lot of nodes…

20$ 1TB egress → 1.5$? that’s -92.5%
1.5$ 1TB storage → 0.75$? that’s -50%

I can’t imagine people still running nodes with such high reduction in payouts.

12 Likes
10

but that would mean the highest Storj could pay SNO’s
is like
1$ pr TBm
5$ pr TB egress

thats something like a 35-40% reduction in earnings, that would make it really hard for this to actually be worthwhile…

maybe in a couple of years when the prices pr TB drops… but then everything will be switching to SSD’s, which is a bit of a wildcard… in all of this…

not sure how we are expected to grow on those levels of payouts, but i guess thats how it always ends with crypto projects…

it’s good while they still have a runway and then when it comes to the actual takeoff, there are no wings to carry it aloft…

what i don’t understand, is why this wasn’t looked into much earlier…
because at these proposed levels of payouts, i doubt the network will grow and thrive.

sure Storj can kick the can down the road, but this issue will have to be dealt with eventually.

i think optimizations are a good suggestion… every bit would help at this point, because from what i can see as even the best possible proposed payouts… thats barely be enough for me to keep going.

3 Likes
11

And with this, I’m out. This almost feels like classic bait and switch. Never was the model you used feasible, now let’s cut payment by at least 75%. This will implode the network. Good luck.

3 Likes
12

SNOs will be leaving even though this is just a proposal. I want to document the current status of the network as of this post.

image
image1840×784 202 KB

12 Likes
13

Lets have a moment of silence for when it does, even though its a bit crowed on the network maybe its for the better cause theres just too many nodes right now.

2 Likes
14

Just to “quote” @digitalfrank from his recent post published at the thread titled “SNO’s please don’t buy anything” (that is currently closed) … I would like to express my point of view and let you guys know that I got a feeling that we are being treated like the “workers” of Bernie Madoff. :blush:

The way the promotion for free accounts was prepared by the Marketing Department and accepted by the CEO of Storj Inc., the way the official press release about the strategy of Storji Inc. is going to be corrected at TC, the way the Community Leader announced “upcoming” changes about which he did not know, the way some links are broken at Storj Inc. website, the way rclone is faster than uplink, the way there is a lack of official containers for major programming languages that can be used by developers, the way there is a lack of official containers for major use cases for final consumers … the rate of change and the size of the spread between our current and proposed payouts. And all this is taking place in a white day. I would like to ask a question: IS THIS BRIGHT? I am smiling a little bit of course again. :blush:

I would like to underline that I do not have all sector experience and data that the management of Storj Inc. has at own disposal, however, I do think that there is a systemic error natively embedded in the strategy of Storji Inc. Instead of building the first class “shared economy ecosystem” and instead taking lessons from such companies as Airbnb and Uber, we are drifting in a direction of a perfect competition and the economic model purely governed by forces of supply and demand where price is not only the competitive advantage but also the only entry barrier. Maybe I am old schooled and such a Web 3.0. model is in front of us all. Who knows? :blush:

However, I have to admit that I am a little bit disappointed. Mostly due to the fact that even though I have not seen any new venture capital rounds being announced recently, I really liked the concept of Storj DCS as a product and I was hoping that a steady work will bring increase in revenues and a bigger market share as well as a joy for us, the fanatical SNOs. :blush:

To sum up, I would like to rise the candidature of Mr. @BrightSilence for the position of Managing Director of The Union of Storj Node Operators. I think that Mr. BrightSilence has a deep understanding of the Storji ecosystem, is deeply involved in the activities of Storj Community and will represent us at a level that is desired during the upcoming talks with the CEO of Storj Inc. So we all could have an appropriate ROI (return on investment) on our selected HDDs. It seems that there is apparently no need to compensate for other factors such as our time.:blush:

23 Likes
15

I, second, third … 10th this … just this… aka +10

2 Likes
16

I think the best way to start innovations first.

  1. Decentrolize Edge servises. We have lot of people here who have 500-1Gbit connections here and static IP with unmetered connections. this move will lower significantly your Edge Service price.
  2. Decentrolize Repear services this will also lower the cost of runing network.

about free tier. for my opinion it can last 3-6 months max. client that not paying for services is not client.

19 Likes
17
2 Likes
18

I find this comment rather insulting . The reason you are making these changes is Storj is losing money and you are trying to make the project profitable.

Since Electricity costs are a now major factor in whether a SNO will be viable now has Storj done any modelling on how this will impact the expected geo location of nodes?

What happens if you sign a large client with Geo location requirements and then the number of nodes dips below the required minimum for stability?

Storj will now move to a hardware subsidy model for some SNO’s rather than any actual profit. By this I mean if an operator makes a decision they want new hardware anyway for some other purpose then the hardware they replace could get moved to Storj to see out any remaining life.

No actual profit at all - just subsidizing the cost of an already planned purchase.

1 Like
19

Why the reduction of payments for node operators? This item of expenditure is only 5% of all costs and is a key one, but they think to reduce it.
As they say in the CIS countries: “Again, these effective managers …” (this means that they are trying to reduce costs where it is not worth doing).

image
image1386×735 106 KB

16 Likes
20

Well You should have start with those line first:

“When we do change SNO payouts, whatever the future amount is, we are planning to roll it out in a gradual way which will include some sort of surge payout while the network growth catches up. We want to avoid any sudden drop in payouts. It will also give us time to evaluate how well it works for the community and respond to ongoing feedback. Given that our initial plan is to transition to the new pricing with surge payouts within the next month or two, then the level of surge pricing will gradually reduce as the demand grows.”

before You showed the proposal rates, coz they are scary.
You better edit that while topic still fresh.

If the traffic load will increase and recompensate new rate to current rate,
and SNO will actually earn the same monthly, OR more,
the rates doesn’t really matters that much.

Just that You said one month or two, wow’s me.
That would mean You expect some really big incease in load very soon.
Coz i would say it would take much longer, because SNO will have to monitor theirs monthly income, and gives You feedback monthly how its going.
And You have to adjust acrodingly, so it seems not so fast process.

Also i dont understand the range
“Egress (per TB) was $20, propose: $1.50 - $5.00”
soo whats Your proposal for us to give feedback is? for egress? $1.50 or $5 or … $3?
if You want to give us $1.5 then give us minimum 13.3333333333 times more egress load,
If You want to give us $5 then give us minimum 4 times more egress load please.

Now, i anticipated it would take us to this situation.

So i proposed STORJ to control new nodes sign up’s, so that most valuable, because
most experienced, established old nodes will get what they need for sustainability.
It didn’t happen.

So now, nodes who shouldn’t been able to join in fist place, joined.
Now EVERY SNO will have to pay the price of that, in form of reduced payout for storage and for bandwidth.

In case of no rise in egress load first, that’s like lose-lose situation.
Hopes its not the case.
(As i said lower You rates, sure, but FIRST You rise the traffic. here)
if it’s hopefully not the case,
which we can’t know,
were kinda scared now(!)
(coz You can’t say, coz customers info joining is confidential)
if it’s not the case,
then we will be fine, no matter what’s the rates per TB are.

In case the traffic doesn’t rise…
Why SNO’s have to pay for YOUR decisions? that You didn’t shut doors to new nodes making,
when overpopulation, oversupply, was obvious even 2 years ago when i posted about it:
here 1

here 2 - “(…)Don’t You understand that its better to have lets say 8000 nodes, and pay them well, than 16000 nodes and everyone kinda unhappy? because it what its goin to right now, especialy the new ones will make BAD NAME to Storj after that unpleasent experience.(…)”

here 3 - “(…)its better to pay all participating people good, and manage/balance inflow of new nodes joining, than let everyone join free and dissapoint everyone with poor payment(…)”

here 4 -

"(…)of coruse the minuses of such managment would be it should be monitored carefully, because its a balance(!!!) beetween:

  1. paying nodes really good enough (i said more FAIR to theirs efforts, its resposible job i have nodes goin offline every couple of days for whatever reasons (for ex: ip operator domain fails) and i have to act)

and

  1. having network capasity large, to take new big customers with no problem.

And the PROS would be as always, the more restricted something is, the more interest it gets. Again i did not invented this, Stroj started ON/OFF’ing new nodes long ago.(…)"

But now, 2 years later,
You still have to do the balancing, just now, in between payments for SNOs,
so i guess balancing was innevitable, just now it will be painfull.

Now OLD nodes will hate You, and new Nodes will hate You.
If population contol scenario was on, today:

  • OLD nodes would be OK, and understanding if cuts would be nessesary,
    (but thats not sure thing, that any drastic changes would be nessesary)
    and
    New nodes would not hate You, because they wouldn’t be there in first place.

So old nodes would be respected by You, so YOU would be respected and understood -
“there is only so much traffic to get at the moment”

At current situation.
Well, i guess we, the old nodes understand that, this way of lowering rates,
it will probbaly comes to the same results.
But in much unpleasant manner.
Old nodes will not be feard, will stay and just get the trafiic from those who will decide to go.
But that’s not guarantee anyone will go.
What if everyone will try to stay in hopes for better traffic?
so it will be worse and worse for everyone in terms of income from STORJ,
without any horizon when it could change.

And if after 5-10-15 months no change for better, then the decionion for leave will be much stronger AND will occure in old nodes.
in such bad scenario, even most loyal will be pissed off of gradual decrase for so long.
So thats where don’t making decion of population control ahead, gets us now.

I guess it is understandable enough, but in case it isn’t, it was like:

  • controling population so there is always food for everyone
    vs
  • letting population grow uncontrollably and at some point just kill half population to feed the other half.

@o1eal
2 years later,
i would like to see how it still might mean im “wrong about the general satisfaction level among SNOs.”
https://forum.storj.io/t/restrain-new-nodes-to-provide-all-nodes-at-least-1tb-egress-mo/14228/37?u=ruskiem

To be clear, I guess in current situation STORJ is finally in face of a growth,
Thats what graph shows, exceleration of overall PB of storage,
so i don’t see any space for closing for new nodes now.
That was to be implemented 2 years ago in phase “in construcion”.
I belive now, STORJ is very operational finnaly, and ready to shine finnally.
Now i think storj should focus on geting traffic.

in case what @NotPaidForBugReport showed, there others things to cut first if so.
Don’t scare Your SNOs with cutting payouts, please :slight_smile:
From now on, lets grow together!

2 Likes