I listened to the twitter spaces presentation and there was mention of changing node ROI…
I assume this would be done by reducing the vetting timeframe.
However, making this change will NOT significantly increase node earnings since during the first 9 months of a node’s life most earnings are retained by storj for the current held- back model.
It is only at month 10 that a node’s earnings start to be noticeable where the SNO retains 100% of those earnings.
If the aim is to significantly change SNO ROI then the held back model would also need to be changed as well or you would literally be mucking around with cents. It has been proposed on the forum many times that the held back period should be increased to greater than 15 months and a lower percentage retained by storj on the early life of the node. Does storj have any updated figures on average node life you can release?
I didn’t use to consider energy costs much, but with prices skyrocketing I now have to consider that running the HDD alone will cost me about 3 EUR a month. That starts to dig into earnings a little. And for that reason I’m currently buying the biggest HDD’s possible when I expand.
I recently bought a 20TB HDD for about 400 EUR. Knowing it would take about 2 years to earn that back taking energy costs into account. It wouldn’t surprise you to hear I use my own earnings calculator to determine that. Of course should payouts drop, this HDD purchase could become iffy. 2 years is already quite a long time.
Personally I have one other consideration that many others don’t have the luxury to do yet. I want to always cover costs of expansion with previous Storj earnings. So if I haven’t made enough money since purchasing my previous expansion, I would probably not buy a new HDD to expand. Now I’ve been around since 2019 and profited from both high surge payouts and testing traffic as well as being able to sell high. So this is currently not my biggest concern. But I think this is where surge payout could help people justify paying for a long term investment with surge payouts.
As for the impact of vetting and held amount. Vetting on the customer facing satellites took about 6 weeks to 2 months on relatively new nodes I launches near the beginning of this year. That has an impact, but on a 2 year timeframe, that really isn’t that important. Held amount is fairly similar and maybe even less impactful. It would only add up to about $26 total and then you get half back in month 16. If neither of these systems existed at all (which is probably not a good idea) it would at most reduce that ROI period by 2-3 months.
However, I’m in the luxury position that I run my nodes on hardware that would have already been online anyway. I only have to consider energy costs of additional HDD’s I run. This is not the case for many node operators who have to take running costs of the additional hardware into account too.
In any case, $5 is a robbery!
Increase the price for end users by $1 for traffic and storage, customers will not scatter, but if operators are paid 4 times less, then everyone will leave.
Keep those $2 off your markup on expenses. Add to your $0.09+2.0 and it will increase your income by 20 times.
I don’t follow your suggestion here. They also lose $13 per TB of egress. $2 more is not going to cover those costs.
But whatever I mentioned was just a shot in the dark. Just a starting point to discuss. But I can’t figure out what you are suggesting without you mentioning all the numbers.
What was the original plan? Pay existing users at the expense of new ones? Pyramid?
At Google, for example, storage of 2tb per month costs $ 5.7. You can’t beat Google! Everything is lost?
You’re making even less sense now… Who’s initial plan? How is it anything like a pyramid scheme?
If you’re wondering about Storj Labs initial plan, they used to charge customers $10 per TB of storage and $45 per TB of egress, making a decent profit. But they realized to break through as an unknown name and a new type of storage, they needed to be price competitive. Customer price dropped to $4/$7. But node payouts didn’t change. Nothing new for start ups to take a loss initially to grow market share.
I don’t know where your pricing for Google storage comes from, but that can at beat be archival storage. Storj provides hot storage instantly accessible at high speeds.
No, I don’t understand what’s going on.
If the original plan was to sell for $45 and pay for $20 why not use this plan? And that’s why, because no one will buy for 45! But the user is offered to pay $ 5 for just allocated space, no need to install additional equipment, just allocate space on your home PC. But at the same time, draconian SLA conditions for operators apply. For some reason, it seems to me that you saw the node itself and its work only at the presentation in power point.
Spend 500-600 on a computer and try to return the investment.
That’s cute. I currently run 13 nodes and have been running nodes since the V3 network started early 2019. I’ve been an active member of this community since then as well and have read the whitepaper and follow all town halls, Twitter spaces and other Storj announcements. I also created several tools for node operators to monitor their nodes, calculate their earnings or predict what to expect. So can we please go back to arguing the substance and put a halt on the ad hominem?
I literally posted my ROI considerations just a few posts up. I know what the costs of running nodes is. I also know it’s a bad idea to spend 500-600 on a computer to run Storj.
I don’t know what pricing you see for Google one, but for me it says 100 per year, if you sign up for a while year otherwise 10 per month. And you pay that amount whether you use the whole space or not. With Storj you only pay for used space. That makes a huge difference.
This was exactly the reason I asked what Storj data suggests on node life. No point suggesting 16 months if the average node life is less than that.
lot of my nodes are 30+ months, and at same time HDDs are 5+ years old
so i dont understand why do you think life of node is only less than 16 months
Because the last time Storj mentioned the figure they had it was 10 months - but this was a significant time ago so looking to see what updates they have on the data.
I think it is because storj node operating started to become popular, and most of nodes was at this time 10 months old. I think it not because HDD-s errors.
Hence why I am asking for an official update. Anything else is just speculation.
Also I have a queston about EGRES it was told that it is rising from 0.7 to 0.9x of storage, as one of the biggest node operators i cant confirm this numbers. I have 230 TB of data and if it would be even 0.7 TB Egres from each TB stored then I would earn over 3k a month, but reality shows other numbers. I shold get only 800 for last month. when i Get home I will share my numbers.
The Storj V3 network has only existed for 3 years, some satellites shorter than that. And since ytou mentioned it was a significant time ago, it may have been closer to two years back then. Was this average a mean or a median? Did it count unvetted or only vetted nodes? Without context it’s hard to say what that 10 months means. I recently started quite a few additional nodes. My average node age might be around 10 months atm. But I didn’t exit any nodes, they are all still running fine and I have no intention to exit them. And my oldest nodes are already over 3 years old.
If you decide to pay money for running a node, I would say you should be planning to run them for at least 3 years. Of course if you start like I did with existing spare space on hardware that is already online. Any income is profit, so that’s a much safer bet to make.
Yeah, that sentence included the caveat “with some multi-petabyte customers”. And it was specifically about data at rest. My own calculations for the earnings estimator are about a tenth of that (per month). So I don’t recognize that number either. Perhaps he was talking about a different timeframe… maybe a year? I don’t know.
My Egres looks like 0.1 only from stored data, so it is big deferens, if it go to 0.9 then my 500 mbit connection will not serve it. with now prices I can easily get Gigabit connection, but with low egress price it will be problematic. I wold like to aks @john where did hi got this numbers about 0.7 and 0.9 Egres per stored data.
In my particular case the economics are even better currently. I still have two nodes left at the moment. Both are running on drives I would otherwise bin. You may recall my thread starting a node on a drive with known bad blocks - well that node has reached month 12 and is still going strong. Bad blocks are not increasing. The other one is running on the drive with 60,000 power on hours. That drive is now failing SMART and has a tonne of bad blocks but Storj continues to work fine. Of course it will die at some point. It is expected. So, I am currently being paid to run Storj on hardware that would otherwise be disposed of. The node that currently fails SMART is 20 months old. However, given we still want to move when possible, we won’t be investing anything further into Storj. It is nice free money currently.
While the discussion around using multiple IP’s is interesting, it should really be a different thread as this is not related to the Storj economic model discussion and has kind of drowned out the original topic now. Lets get on topic, please.