I’m curious what @Th3Van thinks about the change in payouts. How low can he go, and how will it affect his setup.
2 Likes
You can use a python binding, see GitHub - storj-thirdparty/uplink-python: Python bindings for libuplink
You may use a native integration as well: Implementing Nextcloud Using Storj - Storj Docs
1 Like
Keep in mind, customers also have the option to self host an S3 gateway ST, which would avoid these extra costs. Unfortunately with the current pricing system they have no incentive to do so as the used of gateway MT is completely free to them.
3 Likes
Which is crazy. It needs to be the other way around: Incentives for native integrations and usage and ‘penalties’ for using the expensive additional centralized infrastructure that Storj is operating to make customers happy.
2 Likes
You can: Self-hosted S3 Compatible Gateway - Storj DCS Docs
But it’s mean that you need to run something in your infra. Many of developers prefer to use our Storj-hosted S3 Compatible Gateway - Storj DCS Docs instead, even if usage of a native libraries could be more robust: Storj Client Libraries - Storj DCS Docs
This always was our proposal for years. Someone call it “mantra” 
1 Like
(post deleted by author)
(don’t want to participate in this discussion because posts are deleted without notice)
1 Like
You can’t use the spare drive space on HDDs that store other data, your backups or media files, etc. because storj nodes increase the wear of drives and reduce their life. And accessing your personal data will compete with storj access time needs. I won’t put my files to risk for a few $ per month. The serious storagenodes will always use their own drives. Other thing to consider is that with the increase in energy prices, many don’t keep their machines 24/7 running, but only if they run Sorj nodes on them. So I don’t know if there are to many now, who wants to make more than a few bucks, have lots of TB laying around and keeps their PCs running 24/7, and they don’t have to invest in new hardware and more energy costs to run storagenodes. That “mantra” is obsolete.
My biggest “laying around” drives were 2TB videosurveillance WD with 2-3 years of use.
1 Like
Does anyone more knowledgeable than me care to make a proposal in DCS section, to customers, how they feel about a price increase? If they are ok with it knowing that they beneffit from a great decentralised product, that noothers offers?
@all:
How would everybody do with $2 for storage, $2 for egress, $2 for repair, provided that Storj raises prices to customers to $5 for storage and finds a way to reduce expansion factor to 2.5?
why? The internet speed is still less than an access speed to HDD.
Already done:
What is from test satellites is easy to see and understandable to everyone. A lot of test data is stored.
But lately, even, the same traffic from industrial satellites - how can real traffic be so perfectly even, round-the-clock and the same?
I think this is strange and in fact I think that this is done synthetically in order to take a place at the SNO and psychologically cheer up, so I do not want to bring this topic up. This is purely my opinion.
But the fact that test data will be the first to be deleted (after all, Storj Lab pays for them in full itself) and SNO will lose money even without tariff reductions, I think it needs to be announced.
it’s not closed, you may spin-up as much nodes, as you want.
this is one of the reasons, why we always say “use what you have now and what will be online anyway”.
I keep posting different places asking anyone to show me how Storj plans to make a profit and so far nobody has come up with anything. At this point I’m starting to think it’s a lost cause. I don’t think it’s a scam, I just think people don’t know what in the hell their doing like many people in the crypto space.
I see so many projects spending stupid amounts of money trying to build something with no real direction or understanding of the real world. And especially right now, many of them are sinking fast. Storj is not the only one struggling. Worst part of it is, Storj has a product to sell and is STILL blowing through it’s reserves.
2 Likes
This is worse than the official proposal, with 1-5. The big money come from engress. About the expansion factor, the most important thing we have and others don’t is the decentralisation of storage, and the stability of this can only be assured with a good expansion factor. I alyways wondered why is 2.7x and not 3x? You will have triple backup. 2.7x seems a very strange choice.
a bit of numbers
what we are working with…
8 watts consumption for a working HDD 7200 RPM this applied to just about any disk within the last decade or more.
5 to 10 watts pr core again this covers just about any core within the last decade
watt seems to range between 0.10 to 0.50$ pr Kwh
HDD capacity 3TB, 9TB, 18TB nodes would run at about 10000 Watts for a month
thus the raw operational cost of a node is from
1$ to 5$ pr month depending on power costs.
earnings
from my experience with running nodes over 4 or so years earnings have been anywhere from.
5$ pr TB stored with egress to 2.5$ pr TB stored for older nodes, egress drops a lot on older data.
since these are the current prices, that would mean 1.5$ subtracted for TBm earnings.
thus the egress is anywhere from 20% of stored data to 5% depending on node age.
with the new best proposed payouts , that would be 2$ pr TB stored to 1.25$ for older larger nodes.
3TB HDD would earn best case 6$ pr month and 3.75$ if old data.
9TB HDD 18$ pr month and 11.25$ on old data
18TB would earn 23$ assuming that would be old data mostly.
this would mean 3TB to 6TB HDD’s are no longer viable.
9TB would have a net profit of 17$ to 6.25$
a TB of storage costs about 20$ pr TB, so a 9TB HDD would take anywhere from
10.5 months to 28.8 months to pay off the disk value when full…
and this is assuming no other costs exist and at near perfect efficiency…
so basically even a 9TB disk would not be viable either…
it takes time to fill and other costs do exist.
there are also inefficiencies like partitioning which reduces the usable capacity by 5%
StorjLabs recommendation to max go to 90% capacity
extending the time to earn back the TB disk value by approximately 15%
thus the only viable SNOs would be located in areas with cheap power.
and within that subset, it would only be viable if one has the high end spectrum of HDD’s
so the network would be left with a minute amount of SNOs that could even be viable theoretically.
and as was shown here
even the cheapest possible leased hardware on auction, would still be above the 1.25$ expected earnings pr TB of older data, ofc this example is also using 10TB HDD’s which would basically not be viable for storagenode.
we can’t be expected to compete with below datacenter / hardware prices.
datacenters are built to be immensely efficient and even for their best offers it wouldn’t be viable to run storagenodes.
4 Likes
But better than the official proposal with 0,75-1,00
The big loss comes from egress.
Expansion factor is not a backup or replication.
And when your node is full you’re practically doomed.
If people didn’t run dedicated equipment for Storj there would be no Storj network to speak of, plain and simple! I don’t understand why people don’t get this. Very few if anyone at all is using “extra” space they don’t need on their daily drivers in order to make a buck or 2 a month. Maybe that’s how some got started but then they expand.
We humans invented something called farming a long time ago which clearly can be applied to many things… and if there’s financial incentive, the simple FACT IS if it CAN be farmed it WILL be farmed! That’s life, get over it. Maybe you should have asked people to share space to help a good cause out of the kindness of their hearts… but since there IS a financial incentive to running nodes that tells me you already clearly understand that people DON’T do things for free! …generally speaking of course.
So why would Storj expect that to change now with this insult of a proposal?
3 Likes