Raspberry pi is a prototyping device. It’s not mean to be used in production. (People do it anyway and it’s amusing to watch, but it does not change anything).
And if you do make all the precautions to try to make it reliable — actual specced PSU, heatsinks, shielding, ESD protection, etc — you might as well buy an industrial mini pc, with upgradable ram, instead of tying up prototyping hardware in substandard solution it was never intended to be use in.
Well, however you describe it, it is still able to withstand 100% load, and 2 disks won’t load it to 100% all the same.
It is just interesting how people (not someone specific, just an overall observation) get all high and mighty towards RPi setups with reasoning that it will choke from pro setup leaving network, while the problem lies with them abusing data distribution mechanics and creating an unnesessary point of failure, however small.
If you search here a litle, you will find several posts that some rpi nodes are overloaded already.
when you have big ingress and egress on one node, USB on rpi cant transmit lot of small files so fast, then rising ram usage significantly.
Fair point. There are options though when you can hook up sata expansion board to RPi (resulting in SATAII, I believe).
There was a stress test of the network a while ago:
At some point (I do not remember whether this was during the high egress or high ingress) the RPi nodes had problems coping with so much traffic (I assume pretty much every node with data and with free space had very similar traffic).
Sad to read…
I’m SNO since 2019 and i always want to deliver the best for the customer
in terms of speed and reliability.
In order to do so i’m running multiple good suited nodes equipped
with raid configuration, ssd caches, redundant psu and a ups for reliability on some nasty
failsafe connections smallest connection is 1/1gbps highest is 10/10gbps with potential up to 25/25gbps.
Those setups are not longer profitable under this new conditions (even raspy setups aren’t…) and the network will lose many of this “performance nodes” wich builds the backbone of the network. Those nodes doesn’t struggle with stresstests or filewalker and are capable of much more traffic that new customers would generate…
This is a little bit of a heart project for me so i would make some cuts in profit for sure…
But if this turns into a non profit project then i have to quit instantly.
with 25 Gbit connection you will be perfect adge service provider.
Well how very rude ! Where did you get that from… not used is production, that’s the best laugh I’ve had on this thread 
Lets not be hating on the poor little Pi’s now, just because we are going to be loosing a few $$$, back on topic 
A few things to add after going back and re-reading the proposal.
No, a “free market” would be every SNO having the ability to set their own prices. Prices would fluctuate based on factors like healthy competition and supply vs demand. SNO prices would vary based on an individuals infrastructure, efficienty, geographic location, energy costs etc, where people aren’t simply cut out of the race altogether because of the will of the price god. That would be a free market.
It would appear to me that many SNOs including myself are accepting of the idea of reduced payouts, but not nearly to the degree that Storj has proposed. Many have argued that Storj also charge more for the service as well as optimize in other areas as well. I only just learned today that it was SNOs
who apparently petitioned Storj to reduce prices in the first place in an obvious attempt to attract more customers at Storj’s expense without any understanding of how business works. This was an absolutely poor and outright irrisponsible business decision as far as I’m concerned. The way I see it, both Storj as well as SNOs are responsible for the current predicament we’re in now.
In terms of the cost to Storj, the oversupply problem is quite literally a non issue. If your filling nodes with synthetic data then simply stop. You don’t pay for what you don’t use. If people want to have more space ready and able to be filled that’s actually a GOOD thing for Storj and any potential customers, especially large ones that might require that space. This is not complicated.
Yes, the more people show up to eat the pie the smaller piece everyone gets. This is called competition and is to be expected in a “decentralized” and “free” market. Focus on making a bigger f****ing pie, don’t poison the people eating it!
Well this proposal is kind of a contradiction to this statement isn’t it? Storj is simply trying to put the blame on there being to many SNOs instead of their aversion to raise the price back up and admit they were wrong to lower it in the first place. The customer pricing makes it look like Storj doesn’t even take itself seriously.
I thought the whole point is that Storj is paying more for storage than they charge… wouldn’t that be the reason Storj is reducing payouts? Why blame it on an “oversupply” issue that really isn’t an issue… especially for Storj?
Choose to stop? Your forcing them out, plain and simple. This isn’t a charity, and since they have no option here it’s clearly not a “free market” either.
Are you kidding me? How much more efficient can you get than a 5-10 watt SBC and an ~8w spinner? Do you mean for people to go invest in solar now to run these things? Kinda goes against the whole mantra of “don’t buy anything” don’t ya think? And how would that ever turn a profit? The only way for things like this to be more efficient is through scaling! Which is what many are trying to do which apparently you don’t like and call an “OVERSUPPLY PROBLEM”!
The understanding of your business? What part, how cheap highly efficient datacenters can work with? And I’ll bet much of the feedback you got from SNOs willing to cut payouts are probably oness runing their own mini-datacenters which are far more efficient than all the little guys, am I right?
This market doesn’t really fluctuate… it just slowly creeps down over time as data get’s cheaper. So this is just going to continue changing because Storj clearly has no real idea what to do about the huge jam they put themselves in and aren’t willing to readjust customer pricing along with slight cuts to SNOs? Which would of course solve all of this.
So charge the customer for it. This is mainly a convenience factor.
Yes… the customer pricing model!
Sounds to me like you simply didn’t know what you were doing and didn’t account for this cost. Funny enough, you even decreased customer pricing before realizing this… and instead of re-increasing customer pricing because your afraid to loose them, you’ll just take it from SNOs… who you’ve made very clear your not afraid to loose. The same people who’ve been here with you since the beginning helping this whole thing get off the ground… the people who’ve been waiting patiently for years for Storj to take off so everyone can make a more meaningful amount of extra cash for their efforts. But I suppose that’s pretty typical. It’s always the little guys that pay for the big guy’s poor decisions.
Is this Storj’s way of telling us they essentially want to do some heavy stress tests of egress data without incurring the extra cost?
Head up @IsThisOn! @Roberto presumably has to be one of those @CutieePie’s friends who constitute the backbone of the network, presumably sitting on 2 x 8 EiB nominal, with constant snapshots and regular replications to other continents, filling 30TiB just shortly after the vetting period finishes and pulling 0 to 8Kb in ~800us, 8 to 16Kb in ~1.2ms, and almost everything else in about 2.1ms with an army of his arm64 machines, each with 320 cores, 2048GB ram, 200Gbs net and 512 NICs. He is presumably scratching his head how to saturate just one 200 with another army of ultra powerful, energy saving metals located mostly in … Paris 
(just kidding a bit of course).
Would like to point your attention to some of the posts by @hatred. I believe it’s not the first time he is providing a good advice (by the way, please see his post about XFS in another thread, I guess that he might be very right on this). All in all, I have to admit that his advice here (this thread), even though not directly straightforward, also might be top notch. I am not sure if just now is the right timing for KYC or KYP (Know Your Provider), however, I do think that probably the e-mail address used for registration of nodes might be sufficient to constrain the network growth. If indeed this is the scenario that Storj Inc. is currently pursuing. I also like the fact that he is elaborating on the topic of 24 rule that in fact is increasing network latency and decreasing reliability.
I would also like to take this opportunity to point the attention before tomorrow’s meeting with the Management of Storj Inc. (particularly the attention of our representative, Mr. @BrightSilence) to the topic of “elephant movements”. I might be totally wrong [just to quote @IsThisOn], however, personally, I believe that: i) “moving elephants” in general is a wishful thinking and ii) “moving elephants” and aiming at commodity markets in general is even more a wishful thinking. Of course, one may try to do it, however, it will not turn to the favor of storage node operators. And I doubt that in medium to long term it will turn in favor of Storj Inc. as well. For storage node operators, open network will attract new participants, which in turn will increase supply and make renumeration lower. For Storj Inc. which (excuse-moi), is in the business of cutting a margin on every chunk, very short term, it might be advantageous, but in longer timeframes probably not sustainable and will bring increased level of competition and damage to the reputation. It is rather not possible to move the elephant as the elephant will sustain such attempts and possibly will bring down the pricing because the object storage is just a fraction of its revenues. And if it decides to operate at negative pricing territory for sustainable timeframes, the future of the attacker does not look optimistic at all. This was with regard to the offers of such companies as Microsoft, IBM, Oracle and some new kids on the block like Google, AWS and to same extent Backblaze. In my view it might be more advantageous to find large enough target market / market segment and to build competitive advantage there, possibly later from there expanding vertically or horizontally further.
I will be looking with anticipation for tomorrow’s meeting retransmission.
The unfortunate thing for creating a node is that you can literally put anything in as an email and it will accept it and give you the authcode to activate a new node which doesnt help controlling how many nodes are created. It is impossible to tell who is creating them cause you can just spam anything and storj cannot track how many legit nodes exists, An how many are from the same person. Which might be important information if they wanted to legit control and slow down the creation of people just bypassing the /24 and cheating the entire network and getting all data in a centralized location. Which means they have data from all over the world that is supposed to be decentralized but is actually in 1 place which the person uses different IPs from different datacenters and forwards it to there house.
Doing this is actually a negative effect that get created when its geo based and data that is coming from a totally different location then it supposed to be adds latency to the customers.
The other is the free accounts theres over 300k account that all have free 150gigs which if you caculate it they have access to Over 30PB on the network for free if abused.
I was referring to email addresses that are already registered, the email addresses of storage node operators who helped build this network.
Well we used to be limited of how many we could create and forced us to use different emails to start a new node, But now its just a free for all.
In the end though none of it really matters anyway. More free space doesn’t cost anything. Unused space isn’t paid. If anything, it benefits Storj as they can show their customers that they have plenty of free space. How would they onboard new customers without unused space available?
For the /24 issue, the more people that do this the more it just equal out in the end anyway. Won’t really make much difference as the network grows. I would think people who do this though would be incentivized to use connections closest to them anyway or the added latency would hurt them wouldn’t it?
Well that was the point of test data to show how much space there is, But if people are taking advanage of the free space thats storj paying for it not the customers. How do we know how many paying customers there actually is if theres over 30PB of free space to be used for free…What stops people from just creating many free accounts?
Again theres something called /24 issue and when you think about people renting IPs that are IPV4 theres a limited amount so most will take what they can get not always local to them but the cheapest they can get it. Plus if you go with the same datacenter your gonna get IPs on the same /24 for sure. You have people using VPNs that arent always local but they want either a public IP cause there stuck on a gnat and it doesnt mean theres gonna be a datacenter close to them. But datacenter IPs normally are prioritized over residential IPs since Level 2 is faster then resident lines and have higher in and outputs. Latency doesnt really matter that much when the datacenter IPs have super low ping but it being forwarded to a high latency. But it doesnt know that…because its the vps that is forwarding the data to the other IP.
But I thought the test data has stopped flowing? And why would Storj be paying for free space? I don’t get paid for how much I allocate to a node, I get paid for what’s used. How can people take advantage of free space? I can’t just go into my config and change it to 1000 TB and start getting paid for it, it has to fill up. And we’ll never know how many paying customers there are, plain and simple. Storj will know, but they most likely won’t give us that specific info. As for the free accounts that get 150GB free, that needs to be stopped. As for /24, ok so they’ll be in other locations but the latency won’t be effected that much… so what? It all evens out in the end. Honestly, if Storj had a lot more data flowing into the network and everyone wasn’t faced with price cuts I doubt anyone would be complaining. But Storj seems to be trying to blame the price cuts on node operators which is total BS!
Ok then yes I fully agree. I thought you meant nodes since Storj seems to be trying to blame “oversupply” as the reason.
I think the 22M is in tokens isn’t it? For Q4 2022? Average price somewhere around maybe $.40 during that time (just a guestimate) so maybe ~$9M. But I get your point. I would love to know where all that’s going. I would suspect it could be to hedge against the banking crash since many people in the know have known about this coming for some time. One strategy that my wife and I took with our company was to spread out some cash between multiple banks as well as cash reserves… although we also put more into crypto as well. We’ve been warning people of this for a while. That could be one possibility which would actually be a good thing if this banking crash continues. Seen as now 3 or 4 other banks are now under review as others were prior to being seized by the FDIC, it’s anyones guess how bad this might get. Anyway, I would hope their not doing anything “dirty” with it… however it surely wouldn’t be the first in the crypto space… trust no one! lol
Yes it was in tokens but at the exchange price in 2022 was between 1.48 and 48cents So I dont know what the amount was taken.
Yeah I dont trust anyone with crypto… Cause the prices swing so much you dont really know and remeber storj price hit an all time high of 3 dollars at one point. So yeah the prices have changed alot and storj could have easily cashed out before the actually Q4…I havent actually been keeping track of all that I probably should have. Im sure people can figure out where the money actually went. Storj isnt allowed to talk about the market so anything is possible. Theres one thing you should never do on a spending sheet putting something as Other… It raises red flags across the board.
Guys, I would really appreciate if this topic was on the topic - This is about payouts. And for the last 40 messages I just read about /24 limitation and some. I would really appreciate if messages here would be really on the topic, I don’t want to read 50 messages just to read about something else. Please respect others’ time, thank you
Then you should tell everyone to stay on topic, cause in this case It is part of the topic.
