Announcement: Changes to node payout rates as of December 1st 2023 (Open for comment)

After previous reduction people steel adding lot of free space, even faster than customers use it.
So this was the sign of further reductions, so will see now what happens after this reduction of price.

1 Like

What’s up, bro?!

This is a very bold and not substantiated statement - that the network is stable.

  • The last days on the European and Asian satellite there is a historical record of repair traffic.
  • There is no basis in saying that the number of nodes has not decreased due to the huge number of zombie nodes (maggots) located in one /24

Nodes are running in at least 12,357 subnets /24.

I think that we can talk about the number /24 - 12,357, and not about the mythical figure of 22,078 nodes. This is just dust in the eyes.

Also, I would not strongly believe the information about free space on the network - free disk space if it is a dynamically expanding disk cannot be checked and I do not believe in the stated figures of free space on the network.

And here we come to the most important thing - what will happen when a large client suddenly arrives and the space on the network runs out. I understand that some percentage of the storage on the nodes is occupied by garbage that can be deleted.
BUT! Let’s consider the option that the free space has really run out - how in this case will the storj be able to stimulate the addition of space on the nodes by operators? The current payments will not allow you to go and buy a disk with a 5-year warranty - the accounting accounting doesn’t add up.

5 Likes

I completely agree with your assessment Vadim. Particularly since Storj stated one goal they had was to increase node utilization rates. They weren’t able to grow organically fast enough to do this so this was the alternative they came up with.

Vadim, this is not an audited free space on the network, literally at all.


No one is stopping you from writing at least 100 free fake terrabytes on each of your nodes.
And there will be a lot of free space on the network :joy:

It takes non-zero effort to run a node. OK, maybe there are lots of people who don’t mind setting up a node on their always-on NAS or computer with 1TB free space and then not doing anything else with it, I don’t know.
However, for me, it’s like this - I’m OK with doing something for free, but only if it benefits other people who also give/get something for free, so maybe I’ll also get something for free from them.
A good example would be seeding a torrent. I don’t get paid for it, but I also don’t pay for downloading from other people. It’s like a big collaboration.
Another example would be open source projects (if I actually was able to program stuff), like the Linux kernel. Sure, lots of for-profit companies use it, but regular people also use it and multiple people collaborating on developing it makes it better for everyone.

Now, consider what Storj has said in other threads (paraphrased) “our aim is large enterprises that would pay us a lot of money and not require a lot of support”. Well, good enough, Storj is a for-profit company, but in that case I do not want to work for it for free. Just like I avoid self-checkout in a store - I am not working as a cashier for free - either give me a discount or just pay your cashier to do the job.

The 10 cents are precisely why I keep the bottles and cans intact and take them to the deposit machine instead of squashing them and throwing them in a plastic/metal trash can (if at home or close to one) or a regular trash can (if away from home).
At least here you are not allowed to squash the cans and bottles or damage them in any other way if you want your 10 cents back, so the empties take up a lot of space in storage until I get to recycle them.
By the way, at one time, the recyclers whined that the return rate of the deposit items was over 90%. I guess they were expecting people to just pay the 10 cents and not try to get them back.

Back on topic:
My attention on keeping something working is based on how much I get paid. Or rather, I can give you an example from the Storj v2 days.
I ran a few nodes on some hard drives (some I had, some I bought) with no RAID and not much monitoring. I had modified the node software to run better and not just load the IO at 100%. Anyway, I got quite a few tokens for it and I monitored the node servers quire closely. One of them (the biggest) had a habit of locking up once in a while (every few months) so I would reboot it.
Later, the use of the v2 network went down, with v3 being in development. My payouts also went down by a lot and at the time I just lost interest in the nodes. I still would sometimes check them, but at one time, the big node stayed locked up for a few months before I noticed it.

I do not really care about the payout rates on Storj - I care about the total I get. Reduced rates would be OK, if the traffic went up accordingly. As it is now, the rates go down and traffic stays the same or even goes down.

11 Likes

Well, I sound like a broken record, but you can’t outrun economics for long.
previous price adjustment

And that was before Backblaze decided to offer customers 3 times the amount they store as free egress traffic!

There is not enough data in the system and one reason for that is the high price for customers. Customer prices need to go down! It is not competitive at the moment!

This also aged well:
get the basics done first!
Half a year later there is still no Synology documentation…

Remember that Adobe Video partner thing that got hyped in the forum?
Well the page itself is still this
Adobe Partners Storj

3 Likes

Yes I totally agree with you on this point.

I cannot begin to tell you just how much I agree with this. I loathe those self service checkouts for oh so many reasons…

2 Likes

Looking back at my previous calculations here:
(Update Proposal for Storage Node Operators - #446 by IsThisOn)
I think it is time for an update of that calculation.

From a customer’s point of view, they pay 4$/TB storage and 7$/TB download. It doesn’t matter if they download from S3, HTTP or any other edge server or use the native STORJ uplink.

From the STORJ perspective, they pay the nodes 1.5$/TB and need to pay 2.7 nodes. That is a total of 4$/TB they pay for storage and 2$/TB egress. What we don’t know is how much they pay per TB traffic for edge services. We only know how much they pay for edge services in total. 2.7M in Q4 2022 or in other words, more than twice of what they pay to nodes! I assume their costs are roughly the same as the rest of the industry so I will use 5$/TB for traffic. Maybe someone from STORJ can chime in on that.

From a node perspective, we get 1.5$/TB for STORAGE and 2$/TB egress.

What does that mean in reality?

Use case 1: a customer uploads 1TB, stores that for one month, and then download that again over a native integration. In that case, the client pays 11$ (4$ storage plus 7$ download) to STORJ. STORJ pays 6$ (2$ egress + 4$ storage) to nodes. This gives STORJ a profit of 5$

Use case 2: a customer uploads 1TB, stores that for one month, and then download that again over S3. In that case, the client pays 11$ (4$ storage plus 7$ download) to STORJ. STORJ pays 6$ (2$ egress + 4$ storage) to nodes plus 5$ to run the S3 gateway. This is a net zero.

The new economic model finally makes kind of sense.
Too bad that the current customer pricing is not competitive and needs to go down. Which will change that calculation once again.

2 Likes

I made several time suggestions to Storj to make some kind of traffics nodes that will handle gateway traffics, this will lower Storj expenses very much. For Example Hetzner traffic is 20$ per TB

In Germany you can get contracts where you pay 5€ per tb traffic. (Or maybe even lower)

This argument has used so much times, and never confirmed…
This is a usual supply-demand process, if the usage would grow (it’s not instantaneous), SNO will add/expand more space. And yes, it’s possible to delete the remained test data too. Honestly I do not see there is a problem. The only problem what I think is proxy nodes, but I hope it will be solved too.

1 Like

Storj gives a lot of space and egress for free - remember the free tier? And also some non-profit organizations like the Internet archive or documental materials about crimes.

This is what is happening in Germany, if you want to utilize an old car.

You may vote for that question with a heart :slight_smile:

1 Like

While it sucks that the drop for egress is significant (and certainly bigger than I was hoping for), I do recognize that especially for older nodes, the actual payout difference won’t be that big. And I appreciate the promise that payouts are no longer expected to drop further. At least we know what we can count on now.

Payout totals I would guess. Younger nodes will be hit harder by this if looking at the ratio. My older nodes will not feel a big impact at all. I do hope they do something with the held amount to compensate new node operators more, as the impact is felt there much more and they already make so little due to held amount.

As mentioned above, the impact is quite minimal for my payouts. Though I have been reevaluating some of the smaller HDD’s I’ve been running for a while. At this point my intention is to at some point migrate those to larger HDD’s which will have a better cost/income ratio. I don’t plan on stopping them, although I know I’m running some at a loss currently. They were already running at a loss prior to this change. I’m just playing the long game there to avoid having to go through vetting/held amount when larger HDD’s become available to me.

It’s definitely worth for Storj to look into that further as well in addition to the numbers they have provided. Enough nodes, but clustered in a few geographic areas would not be good either, for both performance and reliability.

It’s not greedy to want some compensation for the efforts you put in. But I think small compensation is justified for the reasons you mention. You can still run a profit almost everywhere in the world, if the only cost you have is the power usage of the spinning disk. In my case even enough to buy and additional HDD from time to time. However, at this point you do kind of have to be willing to donate your time setting things up. If I would have been paid my normal salary for the hours I spent on Storj, I would have made a lot more money than just running my nodes. But it’s a learning experience and fun for enthusiasts. And also, I spend much more time on it than I have to to just run the nodes. I helped a friend set up a node on their NAS years ago and they never looked back at it and it has been quietly making money for them in the background on unused resources. I doubt they have even checked their wallet in years and it’s not in their way. So they’s probably be happy to see +/-$200 they didn’t know they had. (It’s a small node)

Well, it hasn’t happened yet. But if an exabyte scale customer comes along, Storj would have no choice but to tell them to look elsewhere right now. Same for the larger petabyte scale customers. It would require quite a bit of organic growth to be able to accept those with open arms with the current strategy. I do think that’s a downside of this approach, but it might be the only way to balance price vs pay at the moment. There’s nothing wrong with growing slow and steady, but that’s what it will be for the foreseeable future as it stands.

4 Likes

So in Fact the „only“ change is the egress-Rate to 2$?

While I have an account, I do not use it, other than sometimes doing a test out of curiosity (when Storj announces that the latency or something was improved). If running a node did not pay any money I would consider doing it if I got a service in return. While I sort-of do, I do not use it. Using a torrent analogy, it’s like downloading and seeding a movie I never plan to watch.

I am running my node and I will continue to do it, since most of the work is already done - the VM runs OK, the hardware runs OK, monitoring works and I only need to update the node once in a while.
However, I don’t think I would create a new node, even if I had another good internet connection or if my current node was disqualified. Very low traffic means it would probably be a year until I saw even $10.

Basically, it something is a collaboration for a common goal, I can do stuff for free. If people rake in profits out of my work, then I want a cut :slight_smile:, which can be in the form of money or services that I use.

]You have to pay a tax to use an old car? They plan something like that here but it is super unpopular, because it is basically a regressive tax.

1 Like

Yes, I think so. And for me personally it has a little impact, but my nodes are old and they still pays my bills, so this is why I want to see an impact on other SNOs

1 Like

Yes, you must and it’s rated based on the age. And also you need to pay for impact on ecology, if your car produces waste (CO² mostly) worse than a standard…, like a diesel.

What Storj Inc. should observe now is not the number of nodes dropping out, but egrees bandwidth. I suspect there will be nodes that instead of dropping out, will just heavily shape traffic against storage nodes.

Oh dear. I haven’t seen plans that bad in my part of the world. For example, the “worst” downlink:uplink ratio we have here is around 10:1.

I don’t think we should optimize the whole network for just one geographic region’s economic constraints though.

It’s, frankly speaking, still a good proposition for me. My operation is, let say, “worth the effort” for around 1.2 USD/TB, maybe even a little less. Right now it is 1.43 USD/TB for me, and after the change maybe 1.35 USD/TB. Good enough, and I’m still ready to add a new HDD for Storj at that price.

Micronodes, I tell you!

And I think Storj Inc. stated it’s not a simple thing to do.

Someone will quickly set up nodes in cheap datacenters like Hetzner’s. I estimated the costs to be below 1 USD/TB to run nodes at scale there. Decentralization won’t be as good, but it will handle the traffic.