Update Proposal for Storage Node Operators

[Ruskiem]

Click to unfold, quotes and replies:

littleskunk, i think this new approach address all 3.

@littleskunk i think this new approach address all 3. here

@littleskunk

There are currently 3 problems:

1. Customer pay just 7$ per TB download traffic but the storage nodes get 20$.
2. Lot’s of free tier accounts that don’t even pay the 7$.
3. Test data that is not getting paid.

The proposal sounds like all 3 points should get addressed. They are all more or less equaly expensive / loss for the company. My personal expecation is that they all 3 need to get addressed otherwise this doesn’t work out.

s-t-o-r-j-user, IT MIGHT be in favor of SNOs, here's how:

@s-t-o-r-j-user IT MIGHT be in favor of SNOs, here’s how: here

@s-t-o-r-j-user

My understanding is that the perfect competition economic model definitely is not in favor of SNOs. Instead of examples like Airbnb and Uber where participants collect rewards that allow them to improve their premises and upgrade their vehicles, we are drifting into a Bitcoin and Chia territory with ever increasing race among the participants and diminishing returns. Even worse, because for Bitcoin and Chia there seems to be some incentive to upgrade hardware to a higher standard but here the incentive seems to be to go not for the second hand hardware but after the proposed changes rather for the third and fourth hand.

BrightSilence, This numbers might ending up with Storj taking over the internet:

@BrightSilence This numbers might ending up with Storj taking over the internet: here
i know the edge service etc, mayby its time for STORj to stop using others, and theirs ways, for marketing leverage, and start being a brand by itself with its OWN solutions of unmatched quality.

@BrightSilence

I’m seeing a lot of people suggest different numbers for payouts, most of them ending up with Storj still losing money.

For context:
Storj gets $4 per TB for storage. Due to the expansion factor (lets assume on average that’s 2.5x) that leaves them with $1.60 per TB of data stored on nodes. They also have to pay for repair processes with this amount.
Storj gets $7 per TB of egress, no expansion factor applies to this as customers pay for all egress used directly. They also need to cover the cost of edge services with this.
And I didn’t even mention the partner program, partners who bring customers on board also receive a cut.

Suggesting payouts higher than $1.60 and $7 is just meaningless as it is not profitable and scale will only make that worse. If you’re going to make these suggestions, please also suggest how they could pay for those higher payout numbers. Otherwise we’re just dreaming out loud… and in that case I’d like $100/$500/$500 please!

Bivvo, This proposal would give You $54/mo for this setup.

@Bivvo This proposal would give You $54/mo for this setup. here

@Bivvo

Tadda, here it is:

@Bivvo

• Icy Box IB-3740-C31
• 2x WD Red PLUS NAS, 10 TB, HDD, 3.5"
• RPi 4B

Average load of 32 Wh = 0.768 kWh per day x 0.40 EUR energy cost = 0.3072 EUR per day * 30 = 9.216 EUR energy cost for 2 HDDs + 2 operating nodes per month.

If the HDDs give up their lifes after 2 years, I would need an income of constantly 33 EUR per month, which is not the case at all already. If they last 3 years, it would be 25 EUR per month, which is also not yet the case for now.

Feel free to copy and edit, folks.

IsThisOn i would like You to look at this proposal

@IsThisOn exactly, “storage has to be as high as possible for nodes not to shutdown” i would like You to look at this proposal, here

@IsThisOn

@nyancodex

$4/mo is dirty cheap.

No it is not. AWS glacier instant retrieval and Backblaze are in the same ballpark. So no, 4$*TB is not cheap, it is a fair average market price for what STORJ is.

@nyancodex

Higher pay for storage like $2.5/TB

I agree that storage has to be as high as possible for nodes not to shutdown because of electricity pricing.
I have read that the factor of STORJ is something like 2.3. So to store 1TB customer data, they need 2,3TB node storage. If we follow your suggestion, that would be 5,75$ per TB. What customer group would pay 5,75$ to STORJ, when they can get Backblaze for 5$ and AWS Glacier for 4$?
If we assume 4$ price / 2,3 = 1,73$ per TB. That is why the current 1,5TB was fair and reasonable.

Now to egress. It is “free” for nodes. Ditch the centralized and unprofitable S3 gateways. 5$ per TB is a fair price for customers and maybe because of that low price, they will overlook the missing S3. Split that 50/50 with nodes, 2,5$ to nodes, and keep 2,5$ for STORJ.

SGC, No, not realy,

@SGC No, not realy, it was just a part of the picture and the approach needed, heres full picture

@SGC

@Pentium100

Maybe Storj made a mistake offering the service for such a low price and including the edge services, which negate the cost-saving “decentralization” aspect of bandwidth.

i clearly remember the days when SNOs laughed at StorjLabs for their high prices and then compared Storj to Backblaze so much so that they convinced StorjLabs to lower their prices.

so we can blame them for being stupid enough to listen to us

just another example in why demagogy is a bad idea lol

also if memory serves the edge services was a very similar deal, people was complaining about not getting enough bandwidth because of the expansion factor.

Knowledge, yeah, but no need to reduce supply, with this approach

@Knowledge No need to reduce supply, with this approach, here

@Knowledge

As some have mentioned, depending on your power costs, it is possible to earn with these new numbers. However, that means nodes have to have data. Part of this, as John mentioned, is to reduce supply. Unprofitable nodes go offline. That’s the only way you really reduce supply. Their lost data will be added to the nodes that remain online.

I don’t know how much synthetic data versus customer data there is. However, I think we can assume that there is quite a bit of synthetic data, and if it is reduced as new customer data is added, we may have flat growth for a while after any initial repairs from unprofitable nodes. It might inspire people to spin up new nodes to capture new customer data, and not lose synthetic data, since only old nodes would suffer the loss/gain while new nodes would gain. If there was a way to prevent synthetic data from being repaired, it might reduce the impact when any nodes go dark.

JDA, i was thinking the same, so i was figuring out a solution for last 2 days, and its this

@JDA i was thinking the same, so i was figuring out a solution for last 2 days, and this what came out, here

@JDA

With the kind of price proposed what you are basically saying is that SNO will be volunteers giving you time, resources and electricity, loosing money in the process and that you hope making a revenue from it?

Trying to get rid of the people that support your business is maybe not the best idea.

PePeR - "Most of people that showed their calculations here, say they’re not gonna bother running nodes." lest change that, here's how

@PePeR “Most of people that showed their calculations here, say they’re not gonna bother running nodes.” lest change that, heres how

@PePeR

@Knowledge

As some have mentioned, depending on your power costs, it is possible to earn with these new numbers.

I could also sell the drive and earn 5 years of Storj income instantly. (Because if I already have the hardware I might as well sell it - instead of playing to be SNO) You say this like the hardware and time is free… Most of people that showed their calculations here, say they’re not gonna bother running nodes. You also mention that if unprofitable nodes will go offline remaining nodes will get better. They won’t… They will get the data, and that’s it. Once they’re full, it doesn’t matter if the network is growing 0x or 20x - people don’t have infinite storage. I think there’s a good chance the network will choke with not enough storage once it grows. You will fill up node in 1 day, and I will wait 5 years to get money what I need to get more storage. I don’t think this ratio is good…

Pentium100, hah i feel You, You got ~ 45TB stored, with this approach You would get ~ $135/mo

@Pentium100 hah i feel You, i see You have got ~ 45TB stored, with this approach i wrote, You would get ~ $135/mo, would like to see Your opinion about it, here

@Pentium100

For February, my node got $52 total, $26.2 was from test satellites (europe-north and saltlake). This is how egress looked:

855kbps (258GB) was from test satellites and 2701kbps (816GB) was from customer satellites

However, stored data was more from the test satellites:

~15TB from test sattelites and ~7.18TB from customer satellites.

So, with the proposed values, I would get $19-$28. If the test data goes away, I would get $6-$11. Now that is not really in the “worth it” level. Yeah, sure, I would get some free space on my server and since the node is running already I may keep it running, but $6 is a bit too low for me to be interested in troubleshooting or trying to keep the node up. It may become like some other things I ran, “let’s check how X is doing, oh, it has crashed a month ago, well, better restart it”.