Vadim? Just imagine! There are people, where money is not a problem. Its Optional!
BTW, do You know safer way to store a .txt file with confidential data, that STORJ?
the text file is small so making it “always above 1000” is no cost at all, and the redundancy is UNMATCHED, it could fuel the STORJ marketing thou the crypto world!
I would like to bump this Topic, because of current payout changes.
According to @Alexey here
If it’s in fact $1.5 * 80 /29 = $4.13, what STORJ pays SNOs for storage of 1TB
then really something need to be done with that part, because it collects $4 from customers.
(i know its redundancy of a file, 80 pieces total /29 pieces need to complete a file,
its a factor of ~2,75)
If You would change a default redundancy factor from 2,75 to lets say 2, You could pay SNOs $3 per 1TB and still cash out $4 from customers. That’s fair 25% for Storj inc.
And we still would be getting $1,5/TB/mo as SNOs.
Additionally as this topic originally meant,
You could offer customizable redundancy for specific files or group of files.
So the client can decide how much safety he wants, and to pay accordingly.
Just so STORJ inc. would always have a healthy 25% to 15% margin of profit from just storage, after paying SNOs.
In this scenario if client wants 2,75 factor, with todays storage price, they would need to pay a SNOs part +25% for Storj inc.:
$1.5 * 80 /29 = $4.13 for SNOs
$4.13 * 1,25 = 5,16$ per 1TB/mo for storage from customer total.
And if he wants 2 times better than now
$1.5 * 160 /29 = $8,27 for SNOs
$8,27 * 1,25 = $10,34 per 1TB/mo
You could make it even stepwise.
The more redundancy the customer wants, the cheaper it can gets, like:
$8,27 * 1,20 = $9,93 per 1TB/mo
And the fun example, “always above 1000” for Vadim:
A 1000 parts would need factor of 12,5. (if 80 parts is factor of 2,75)
$1.5 * 1000 /29 = $51,72 for SNOs
$51,72 * 1,20 = $62 per 1TB/mo storage from customer
So still cheaper than BIG CLOUD,
if counting storage + egress, according to STORJ graph here, right?!
Because
$62 storage + $7 egress of 1TB, $69 STORJ
vs
$23 storage + 80 egress of 1TB from $103 BIG CLOUD
And they have nothing near the redundancy of 34 like in this case!
…and World WIDE!
P.S.
Alexey said also:
“And this is without repair costs (Storj paying to SNO $10/TB of egress repair, if the repair job got triggered).”
Yes, but that cost goes from the pocket of a SNO, who lost piece, or got offline too long so its always covered.
We are talking here about increasing amount of pieces in the same model, still 80 pieces of one file.
Right now we have around 22000 nodes.
How about i want a piece on ALL of them?
$1.5 * 22000 / 29 = $1138 for SNOs
$1138 * 1,15 = $1308/TB/mo from a customer for keeping 1TB in 22000 pieces WOW!
Assuming that our important .txt is like 16KB, its just 0.000000016 TB
So 0.000000016 x $1308 = $0,000020928 per month for to store a 16KB file on 22000 nodes! WOW, who can do that, but STORJ?!
Who should do that BUT STORJ!?
Also this mechanism, can be very desirable, if someone is preparing to distribute a file to the world.
Like world premiere, or publishing it on popular medium, when the download is EXTREME at the moment of publishing, so the file should be at many many nodes to assure it can be accessed for everyone at once.
Or say i have 100MB and want it’s piece on all 22000 nodes.
it’s 0.0001024TB * $1308 = $0,1339392 per mo to store a 100MB file on 22000 nodes.
Vadim, so its NOT so much. Not much at ALL, for the benefit it gives.
So that’s with still $1,5 per 1TB for SNOs.
END of main Topic
But now lets back to the first part of this post.
If it’s in fact $1.5 * 80 /29 = $4.13, what STORJ pays SNOs for storage of 1TB
then really something need to be done with that part, to not lose money.
… or price for customers should be more for storage of 1TB. (not $4 like now)
Let me show You some simulation:
I would start from the foundation of the network, the nodes.
To secure the foundations, i assume you have to pay the node well.
Well is $2-3 for storage of 1 TB (not current $1,5), (and money for egress is included in it) Because paying such high amount to SNO is ONE and ONLY guarant that he will want to stay in network, and allows Storj inc. to drop any held amount, or drastically reduce it (just always 1-2 month delay upfornt of payment). So the new node, can benefit right away from joing the network, and it enables SMALL SNOs to join because of that (now that’s not profitable for small), and small SNOs GREATLY increase decentralization, and this is what Storj needs, more nodes in country’s like india, or asia in general. That is only to achieve by giving them the GREAT offer to be a SNO. (Remember after some years they have to buy some used HDDs, to replace, if want to continue to be a SNO, HDDs are not eternal.)
Then, if everything stays the same, just SNOs pay up, from $1,5 to $3 for storage.
$3 * 80 /29 = $8,27
$8,27 * 1,25 = $10,34 from customers per 1TB/mo, storage.
Looks like “no way, too much!”
Would that be really too much? well if SNO would be paid that high,
as a SNO, i argued few times, that this is really ENOUGH and satisfying!
That means, no need for money from egress.
As a 7TB node operator, i used to be paid $20-26 for it.
For 06.2023, my payout was $13,49
for 05.2023 was $13,68
04.2023 was $16.40
03.2023 was $15.90
02.2023 was $16.12
01.2023 was $18.32
12.2022 was $22.23
and months before about 20-26$ for this full 7TB node.
and around 50% of that was egress back then, now there’s almost no egress.
That’s not stable for a node operator, as You can see, no egress traffic makes 50% difference for SNO payout, also You reduces pay rates for future.
i think what happened it’s You cut synthetic egress and prepared a network to be able to scale, that’s good. BUT. As You can see it results in 50% drop of payouts for a node, and SNOs in general. And with current rate for customers at $7/TB for egress, seems they just don’t want to use it at this rate! (because You cut synthetic, and what’s left is almost none)
Means its still too expensive to use egress at STORJ somehow. (that’s the market, not me!)
Therefore, look,
if You get $10 from customers for storage, You could offer egress for $1 or $2 per TB.
And keep all that for STORJ inc. just like now, where its $7 egress and You keeps $2 as well.
The difference is You would pay SNOs $3 for storage, which allows You to ADD nodes to the network world wide (not lose like now, with rate at $1,5), ADD nodes, especially in india, AND keep the nodes filling, because so small egress price would finally incentivize a real USAGE, especially for Videos, which is according to Livepeer: 80% of internet’s traffic.
And thats before polishing.
How about SNOs would be just as happy with $2,5?
$2,5 is still much better than $1,5.
It only matters what is the real rate, at which, the nodes would keep comming and adding HDDs for STORJ. That need to be veryfied by the market.
if $2,5
that would do
$2,5 * 80 /29 = $6,89
$6,89 * 1,25 = $8,61 from customers per 1TB/mo, storage.
now You could definitely ask for $2 or even $3 per egress from customers.
(currenly its $4 storage +$7 egress = $11)
New ratio would be $8,61 or $8,49 for better look, + $3 = 11,49$
or $8,49 + $2,5 = $11
Still same total outcome nominally.
But what gives better opportunity to profit?
What people want more? storage or cheap usage - egress?
I think people care less about storage price, and crave a real cheap usage option, because that what’s takes gross % of cost.
You upload a video once, and people can download it potentially limitless times.
But You can’t upload limitless data to profit from storage!
Therefore i think it is better to cut profit % from what can be limitless.
And offer cheap egress to profit from its usage.
And to profit from egress usage, it has to be in good price, so people want to use it here.
I think with that approach STORJ could attract video intensive traffic customers.
And at the same time, secure its future, by making sure the network is ever growing, with loyal nodes that would never want to quit.
Thank You for reading. @john
And if someone is afraid people will game it, like serving no egress availability for the STORJ, then i say yes, some people are trying to game everything. But just like in games, there’s just anti-cheating mechanisms to assure the agreement is being honored by both sides. No exception here.