Current situation with profitability per TB is 0.18-0.81 $/TB

[AiS1972]

Greetings gentlemen!

Now there are a lot of messages concerning, in general, the same thing - a discrepancy between the occupied space for the storage data on the operator’s disks and what the storj pays for.
It seemed to me that from the huge number of these messages Alexey was already starting to burn out, and this is bad for all of us.

SNO - I have 10 TB occupied, the storj paid for 7 TB.
Storj is garbage on your nodes; storj does not pay for garbage. And yes, leave at least 10% of the space free, no matter if your node is 60TB in size.
And so on and so forth…

So… These are all lyrics that are not productive for anyone.
I suggest that we all simply calculate the net income and if it suits us, then leave the nodes, if not, delete them.

I took the liberty of patching Krey’s script by adding two variables to it - the total expenses for nodes per month and the total volume of disks allocated for nodes.
This made it possible to get dirty income - which shows the original Krey script and the clean one that the patch gave.

I will look at all this using the example of an American 60TB node located in Wisconsin, USA.

The current situation of this node in the chart

Let’s assume that this is a system consuming 80 watts - 50 motherboard + 30 - 3 20TB disks. This is ideal, in reality we will add another 20 watts for the fact that the Internet has entered the system and the efficiency of the power supplies - a total of 100 watts for everything at minimum wages.
The cost of electricity in Wisconsin is 16 cents - a total of 0.1 kW * 720 = 72 kW * 0.16 = USD11.52 - we consider this all our expenses, we will not even count the Internet, which also costs money.
Our total volume - the number written on the disks - 3 x 20TB = 60TB. And this is an ideal situation. For example, my average disk size is smaller now - 18.08.
Node settings for the Krey script

Description of the two added variables

Script results:

Profitability in the considered case - netto 0,32$/TB; gross 1,21$/TB
If this node were located in California https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_5_6_a

Electricity costs would double to, say, usd20.
Profitability - netto 0,18$/TB; gross 1,21$/TB

In my case the yield is as follows- netto 0,81$/TB; gross 1,17$/TB
This is the average disk size of 18.08TB and electricity for 0.06USD, Internet 1Gbit - usd26 and IP address (average price usd1.5 per IP ) are also taken into account, on average I get 1 watt from the outlet per TB of disks. This is a net yield higher than I have yet seen.

I could consider our beloved Th3Van as an example, but I could not understand the cost of electricity in Denmark.

Let’s sum it up - if someone does not agree with the profitability of 0.18-0.81 $/TB, it is better for them to delete the storj nodes.

For me, debit and credit absolutely do not match for American nodes, as well as for European ones, but oh well - this is all just for fun

Below is the script itself and an example configuration file:
Storj3Monitor.ps1 - https://cloud.mail.ru/public/1ou1/butZRyNEL
test.conf https://cloud.mail.ru/public/xaRq/GK3H9m3YV
run script “C:\Program Files\PowerShell\7\pwsh.exe” -File storj3monitor.ps1 -c test.conf -TimeoutSec 300 -p all

Throw in your netto profitability for comparison!

Thank you for your attention.
See you again!

[Roxor]

So after including a variety of expenses: your examples are still making money from unused capacity? Excellent! Must be why we have 24k+ nodes

[Copycat]

Привет.
Видел изначально пост на русском, отвечу на русском.
Спасибо за проделанную работу, и за изложенные мысли, это интересно.
Сторж изначально себя позиционирует очень правильно по поводу доходов: используйте то, что и так используется, поэтому если хоть цент заработан - хорошо.
Но я понимаю СНОв сам такой, тоже неприятно, что выплаты упали.
К сожалению для СНОв чиа показала, что владельцы дисков готовы их эксплуатировать даже если доход от этого едва будет перекрывать затраты на эл/энергию, так почему этим не воспользоваться. Бизнес, ничего личного. Тут я бы так же сделал.

Мне проект изначально очень сильно зашел благодаря техподдержке, Алексею огромное уважение за работу.

П.С. помню как первый раз увидел датского гиганта , сейчас как-то грустно смотреть его статистику.

[Vadim]

What about my setup, I dont like to assume, I will tell in the beginning of next month by real results.
Today i have 359 TB, it was 400 before delete start.

[AiS1972]

@Copycat

Видел изначально пост на русском, отвечу на русском.
Спасибо за проделанную работу, и за изложенные мысли, это интересно.

Весь смысл этого поста - поставить сно текущего/планирующего им стать, перед нетто доходом. Потому как на данный момент $1.5/TB - это скорость коня в сферическом вакууме, потому корректно там где 1.5 , написать в скобочках (netto 0.1-0.8$/TB, depends on hosting region).
Это уберет все лишние и ненужные вопросы - ой а мне мерещится что у меня занято 10 ТБ, а сторж заплатил за 7. Пожалеем Алексея, он у нас один.

[AiS1972]

@Vadim

What about my setup, I dont like to assume, I will tell in the beginning of next month by real results.

Vadim - it doesn’t matter how much data storage you have, what matters is how much is allocated.
Therefore, it is easy to calculate the netto -
(payment for March - all current costs) / volume of allocated disks.

[Vadim]

My calculations are other, I see how much i got to my account-how much i speeded- then i know is it profitable or not. Allocation i changing all the time, for example tomorrow will start move additional traffic, and I will get 1-2 TB of data every day. Then i will get more money for it. So i am not assuming, I calculate on real results.
For example For last month I got 540$x0.92 - 750KWHx0.17€(electricity)- 20€(internet)-20€(rent)=329€ profit.
It is very simple math.

[pangolin]

@Vadim

20€(rent)

Rent what? Server space can not be that cheap…

[Vadim]

it is small space that i rent for my servers, to diversify location and internet connection.

[pangolin]

What country is this where the main storage costs nothing?

[Vadim]

most of my HDD are used and I got them for free from my work, and other are got from income from storj.

[AiS1972]

@Vadim

For example For last month I got 540$x0.92 - 750KWHx0.17€(electricity)- 20€(internet)-20€(rent)=329€ profit.

sorry, bro…
20€(internet)+20€(rent) - Is this all spending on the Internet and IP?
How many unique subnets /24 do you have and what is the average cost of one IP address ?

[Vadim]

ow yes you are right I forgot about it, but it not very much and i do not disclose this information.

[Toyoo]

Yep, that’s true, if you have oversized hardware taking more than 1 W per terabyte of stored data, there’s probably no point in hosting a node. 50 W for just a mobo hosting only non-commercial Storj data? Way too much!

Consider instead getting some sort of an Arm-based platform, or an energy-efficient x86 platform, these are already capable of hosting dozens of terabytes of nodes. I’m running mine on N36L driving ~80 TB worth of data (both Storj and personal data) at ~50 W at the wall total. I’m planning to expand it with two more 20TB drives in future.

I’m paying 0.21 USD / kWh, and about 13 USD for the Internet connection (which I would need anyway for personal/remote work purposes, so it’s not like I’d count that as Storj cost).

[AiS1972]

@Toyoo

I’m paying 0.21 USD / W, and about 13 USD for the Internet connection (which I would need anyway for personal/remote work purposes, so it’s not like I’d count that as Storj cost).

This is good, but the data is not complete.
There are storage and costs. How much did your nodes earn in March?

[Toyoo]

(spotted a wrong unit in my post, corrected)

That would be 93 USD.

[AiS1972]

@Toyoo

Consider instead getting some sort of an Arm-based platform, or an energy-efficient x86 platform, these are already capable of hosting dozens of terabytes of nodes. I’m running mine on N36L driving ~80 TB worth of data (both Storj and personal data) at ~50 W at the wall total. I’m planning to expand it with two more 20TB drives in future.

Bro!
If you already have 80 TB - that’s 4 disks of 20TB each, will you put two more disks in another case?
You forgot to clarify whether you have an additional fee for IP addresses, since 80 TB is most likely a minimum of 4 nodes

[Toyoo]

I’ve already got 5 drives inside: four in the regular slots, and the fifth one in the top 5.25" slot. The fifth one is very small though.

It is technically possible to fit 2×3.5" in the top slot, there’s enough space inside. But cooling that is very difficult. So the sixth one will go on USB.

I didn’t say I’ve got 80 TB of Storj data. But yeah, I used to have nodes on more networks, back when I was working at a different company where I could add my own hardware to their server rack. Consolidated them in this box at some point. So, no additional fees, just an ad-hoc opportunity.

[IsThisOn]

@AiS1972

Let’s assume that this is a system consuming 80 watts - 50 motherboard + 30 - 3 20TB disks. This is ideal, in reality we will add another 20 watts for the fact that the Internet has entered the system and the efficiency of the power supplies - a total of 100 watts for everything at minimum wages.

I have 0w for 20TB I don’t use.
You know, the only use unused resources mantra? The thing STORJ employees keep repeating over and over in the forum?

@AiS1972

we will not even count the Internet, which also costs money

My internet is 10GBit. Most of the time I use 15mbit to stream Netflix. Sometimes 10GBit to download an ISO. Anyway, the 10mbit STORJ does not really bother me. So yeah, internet should be free.

@AiS1972

Let’s sum it up - if someone does not agree with the profitability of 0.18-0.81 $/TB, it is better for them to delete the storj nodes.

Totally agree. This seems like a fair number!

Bigger problem I see is this:
Performance of STORJ is bad. It is so bad, it basically asks for some kind of cache mechanic.
I for one only trust ZFS. I also only have unused resources on ZFS. I know this is a me problem, but it still applies.

Now with the current low ingress, I really ask myself the question, should I really put the fragmentation on my ZFS pool, just to have a 5TB pool after one year, that then gets me 10$ a month in some token where I pay huge fees to fiat offramp, have to wait months to make the minimum payout and hold back amount, all of that in the hope that STORJ does not go belly up in the meantime?

For me, this is currently a big no. Maybe when we weeded out the free tier abusers, the none unused storage nodes, we discontinued S3, there are no multiple IP nodes, no VPNs, the list goes on. Basically, as soon as I can see STORJ as a working business model, I will join again. And I have nothing to loose, if I join a year from now, instead of now. Either the model works, and then it will also work a year later or it doesn’t.

[pangolin]

@IsThisOn

Either the model works, and then it will also work a year later or it doesn’t.

Storj already gave up on small customers. Otherwise they wouldn’t have closed the free tier offer the way they did.

The next thing to discontinue will be the small SNOs. Why? Because Storj can never meet big customers speed and certification criteria with such home user storage.