Hold on your horses my friend. Those numbers should not be destined for our eyes. And please do stop being such a nice guy, we all might get influenced. 
At the risk of exposing my methods, you catch more flies with honey than with vinegar. 
Besides, I wasn’t asking for anything they can’t share and with a community that has clearly shown to be willing to think along on how costs can be mitigated, it may just be to their advantage to share. In my experience they are very transparant where they can be. So it can’t hurt to ask either way.
4 Likes
Still, this is a public place, thus I would be very cautious. It is always useful to ask a clever question and to receive a clever reply, not always in a public place. One of the reasons I raised your candidacy for the position of the Union’s Managing Director. Keeping such stakeholders as storage node operators informed is part of the game.
BrightSilence you are correct. Synthetic load = test data = data uploaded from non production (non customer facing) satellites by Storj.
3 Likes
Always happy to exchange ideas, I have to say “excusez-moi” and try to somehow trim my logs, until next time.
The remaining 21% is almost entirely the satellite operations costs, but for the infrastructure and the Dev Ops people. Customer support is also in COGS but it is small compared to the other costs.
7 Likes
Thank you all for the time you have taken to read and comment on this post. I have noticed many of you sharing data about your costs of acquiring and operating your nodes. I’d love to collect as much of this data as possible in a form that will allow us to do some analysis. As such I’ve created a survey to capture this data in a usable format.
If you would be willing to fill out this survey and share your data that would be much appreciated.
9 Likes
Done. Though I have many nodes with more than one IP. I have just added everything together. Hope that’s ok. I have made purchases in the past that I likely wouldn’t have if the proposed payouts were in place. I had to get a little creative on the question how long this node hardware could still last. Some of it could last much longer than other parts. Some nodes run on a RAID so won’t fail if an HDD fails and some run on single individual HDD’s. I also keep continuously upgrading and replacing stuff. So I just went with 5 years for now. But I think node hardware should be profitable in far less time than that if you want to be able to still attract node operators.
4 Likes
Thank you, also for this marketing form. Please kindly bring some profitable clients. Cheers to the Sales Departments. Thank you all for the conversation.
I filled the form as well. My “estimated life remaining” was mostly based on when do I think my node will fill up and I will need to buy more drives, because I have no idea how long the hardware is going to last. My hardware is old, but I use an even older server with no problems.
same here, lot of my hdds are about 7-10 years old.
I think Storj should send out an email to all SNOs and give a link to this forum thread in it. There are still a lot of SNOs that don’t use the forum. Alternatively a banner at the top of the dashboard would be good too as many SNOs did use fake/temporary email address during their node setup.
2 Likes
so many post pointing out that this proposal is trash.
Storj lives from Node Operators. Node Operators make sense if they make any revenue, and by “any” i mean it must provide you some worth doing it.
With this proposal the math doesnt check out. You make surveys collecting data although you have all the data you need. Make your research. Go find the Country with the highest electricity costs, go find your average node capacity and speed and make your calculations and tell me if it would be worth it for you. Most people told you “its not”.
This will be a death sentence. selfishness of companies strikes again. The Node Operators are not happy with this.
4 Likes
According to proposal that you made, i can 100% sure to say that im dum idiot spending ~500USD on UPS solution for my node fighting for 100% online score. Also, one more 16tb drive. I would better donate it to Ukrainian army…
According to “new” payouts that i will have, it will take more then half year just to payout UPS, and i am not even talking about hdds (spends are not included).
I have several 26 mouths old nodes, each earned ~150USD (they located in different places). So they do not even close to pay them off for itselves, with current payouts, and i think it will take 1 more year to. 3.5 year ROI looks crazy for such type of investment. Especialy that according to statictics after 4 years i will have chance 25+% that those hdds will fail.
Of course i have hdds thats paid their price x2 in last 4 years. Buuuut, even if i will stay here i wont invest any more cent in it.
My current month payment around 250 USD for 16 nodes, if it will be 1/4 out of current value i do not think that it has any sense for me to maintain them cause of:
44 USD internet bills on 4 locations, 1 location free for me.
~60 USD electricity bills on 2 locations(probably will be higher x2 in next months due to new laws), 3 locations are free for me.
so as you can see, after you will enable new payout scheme it will become totally not profitable for me to run it.
So after new payouts will be enabled, dear SNOs for whom it will be still profitable, be ready to get additional 65.5 TB of repair traffic)
There was to much told about stupidity to decrease payouts for those on whom you spend only 5% out of all your payments. Taking in account that those 5% - are your storage providers.
Just think where storj will be without them?).
And about edge services, why should SNO pay for this, and not those who need it, why you can make it as additional payed option for your customers?
About my thoughts where storj need to move to take more profit - individual customers, to compete with icloud/google photos etc, because storj has huge benefits against them. With good marketing it looks promicing especially if you will have certificates.
As for myself i would like to have some module for nextcloud, that will allow me to store my data securely without any chances them to be lost due to some hardware malfunction, and so on. But it look like just a dreams, because individual customer are not interesting for storj…
p.s. filled survey just for the first my node, cause it will be annoying to fill it for all 16 of them individually…
p.s.s i would still advice my company to choose storj to keep our logs (around 30gb per day), because it looks dirty cheap in comparacing to azure or amazon s3 solutions, but i think we will wait at least few months after payout scheme changes to check what will happen with network.
1 Like
This is getting out of hand…
- You need UPS for data integrity, not uptime. If you did not have UPS before — congratulations, you have one now, should have done it years ago.
- There is no requirement for 100% uptime for a storagenode.
If you did not need a drive for other things — why did you buy it?
Leave politics out of this.
You really need to read the thread or at least listen the twitter space once again, you misunderstood. Also calling decisions you don’t know all facts about stupid is questionable.
Assuming you need to store logs for a long time for archival purposes and expect to never or rarely need them — glacier deep archive is much cheaper than storj.
2 Likes
- Goal was to be reliable node, as i as i part of this network, and i do not want my node to be a problem to it. Earlier i wasnt need ups that will be able to hold nodes online for 8+ hours… But situation changes.
is there somewhere hidden requirment tells that i can be offline for a half of month?
Hmmmm… my nodes were almost full. why not while it is profitable enough and im ok with ROI that i will have?
Where did you saw me talking about politics?
The this happening in Ukraine is called war. Learn to call a spade a spade
i have already spent to much time checking 200+ posts, so im not sure what i can find interesting in other ones.
glacier deep archive - its not an option, cause data need to be highly accessible in any time.
nevermind…
The point i was talking about is that with such low rewards, it is not economically viable to even keep node online.
According to storj expectations that there will be some peoples, who will share part of their spare space to network is too naive. Just imagine situation, i have pc that is running almost whole day, i have for example… 8TB spare, i’m installing storj, and now i need to run my pc 24/7 (or i will have low online and after some time i will disqualified).
What sense to me run my pc for extra time, share my hdd with storj, that will use it under its limit for almost all time, it will have more chances that it will die faster and so on. Also i wont be able to quickly change amount of space dedicated to storj if would suddenly need to download some huge files. What my interest to be limited somehow in use of my equipment, spend time to setup/maintain node for such low price? Taking an account that it even wont payoff electricity bill, and im not even talking about internet connection.
And dont tell me about NASs… try to watch at least 2k movie from hdd that has 150+ iops…
This concept won’t work. He is dead, no one will do this out of ideological / good motives (ok, almost no one, but thats definitely wont be enough).
Until peoples will see economical interest to launch/run/maintain nodes.
For 200+ posts i have already saw ~10 SNO who told that they will leave right after this poor decision will be approved.
Yeah 10 SNOs for such amount of nodes is almost nothing, but im pretty sure that there will be much much bigger amount.
Will see what will be next…
1 Like
Never been an issue for me. And I frequently watch 4K movies from my NAS which has 14 nodes running on it as well, 5 of which run on the same array my media is on.
3 Likes
Sadly, I expect Storj Labs keeps selling STORJ token reserves at the same level than previous months (around 30M STORJ in 3 months).
Taking into account that current reserves are around 120M STORJ, EOL of the project could be in around 1-1.5 years. Monthly payout discussion won’t be necessary anymore.
2 Likes
When people say UPS by default it’s understood that it’s a short runtime power delivery to ensure safe server shutdown when power is lost and to tolerate occasional brownouts.
If you are taking about adding reserve battery or generator system to maintain uptime just for storj — it makes no sense. It will never be worth it. You would then need second ISP line, hotspot fallback, all nine yards. It is unreasonable.
If you already had all of that — then sure, let storagenode use it for free.
Nobody can maintain 100% up time. Not even Amazon. What that marketing high level page means is “keep it always on, on a best effort basis”. If you read actual details docs you’ll see that few (six? Ten?) hours per month is harmless, and much longer is outages are also well tolerated and recoverable.
Back to UPS discussion: the conventional meaning is to protect against data loss on power failure, runtime measures in minutes. Therefore adding UPS to improve runtime makes no sense: you can have few hours long outage, reducing it by 5 min makes no difference. It is however absolutely necessary for ensuring data integrity.
I guess you have a point here: if you already have the whole server for other purposes and have an empty bay and can source cheap used drive on eBay it may actually make sense. No objections here.
Right. Building a storage node would not make sense. You can’t build a competitive one-off hardware, even before accounting expansion factor, and compete with hyperscalers.
Carving out unused on or adding space to the existing system — totally can be viable, because everything else — already exists and is free. You are essentially just paying for an hdd and 10W of power, not the whole server.
Then this system is not suitable for storj. My server was already running 24/7, and I had free space so I addded storj. Running It cost me $0.
That’s not a thing. Drives are design to literally run all the time. Using them as intended does not shorten lifespan.
This is a valid concern. You can start many small nodes, and evict nodes when you need space.
Movies are sequential reads. Correctly configured nas should have no issue running storj, scrub, and streaming at the same time. Mine does, as does other folkses.
This is not representative of anything. Anecdotally I won’t leave even if payout is 0. It costs me nothing, why not donate extra resources? I run my server for other reasons anyway, and have unlimited internet bandwidth. So, why would I shut down perfectly working zero maintenance node? If it pays anything — it’s beer money :).
Yep. Super curious. But as I hinted above, i don’t think it is correct to expect for node payouts to pay for the hardware in full let alone profit on top: hyperscalers do it on scale, and don’t have 2.7 amplification factor. Storj can only exist as an environmentally friendly service utilizing spare capacity, just as it was conceived.
6 Likes