Costs of running storagenode

Has all of the downtime requirements been run through a model using current stats to show what the impact would be to nodes? Would it be one percent, five percent, ect?

Some of the things thrown out in this thread what SNO should be doing for uptime are crazy unless people just really want to donate time\money\other resources to the project. There is no way people can even come close to breaking even running an HA environment with current payout amounts IF you factor in all your costs. People that say they are making money might be if you leave out key factors that you really are paying for but in reality are just donating to the project.

Some things are indeed crazy to do for a node. You shouldn’t be doing anything more than you would for your personal NAS. In fact if you already have one, use its spare space to run a node. That’s what I do. No extra costs, pure profit. Ok… I’ve bought new HDDs to expand my array for storj now, but I’ve done that with my storj earnings, so I’m fine with spending what I already earned. And I’ve never been in the red on this entire endeavor. Just don’t go overboard with your setup. People are running successful nodes on nothing more than a raspberry pi.

1 Like

Great example. That personal NAS is going to have be replaced eventually. Are you factoring in the expense of that? Are you factoring in the expense of your internet connection? Are you factoring in your time?

I am part of the project because I find it interesting and have learned a few things along the way. Lets call it like it is. If you factor in all those things payment would have be way different to even come close covering expenses. You are donating to the project which is ok, but lets be real about it. You are donating part of your home setup to be part of greater community that you get a few dollars back.

Your already paying for internet so you can’t include it so other wise if you werent running a node you would still be paying for internet.

If you enjoy doing it your time isnt a factor either. I waste alot of time doing alot of things that I dont get paid to do.

1 Like

Incorrect, you have to start factoring that in if you really want to be real about what you are spending. I would guess of you go read your TOS of your home connection that running it for business purposes is prohibited.

No you don’t if im already paying for internet at a rate already adding a node to my internet doesnt increase the amount so no you do not have to include it. If I added another internet connection solely for a node yes you would include it. Otherwise no

This has actually nothing to do with downtime disqualification blueprint. Discuss it somewhere appropriate please. (But I guess @Alexey will just move the whole discussion to a new thread)


We can agree to disagree. If you are taking resources that you are already paying for and using it for a purpose there is a cost you have to associate with it. You are choosing to look at it differently and not a pure what are my real costs when it comes to this.

Sunk costs are not the same as opportunity cost

I agree with you on time use, but not Internet connection costs if that connection is unmetered. However, if one has a metered Internet connection, then the excess bandwidth would no longer be a sunk cost.

1 Like

Yeah, that makes no sense at all. I’m not factoring those costs in because as I mentioned I had already made those costs and would make them whether I run a node or not.

As for my time, I spend exactly as much time on it as I want to. And I would pretty much have the option to spend no time at all on it. It just does it’s thing in the background and a friend of mine is running his node without even looking at it for months now. I spend time here and on building tools because I enjoy it, not because running a node requires me to spend time on it.


It makes complete sense. If you are not factoring that in your then donating time\resources. There is nothing wrong with that but lets be honest about it when we are talking about making a profit and not making a profit. There is a big difference.

So far my investment as an SNO has allowed me to break even considering that I only added a new HDD to my 4-bay HP microserver for Storj purposes. I guess what I have earned so far in Storj since the last 7 months has written off the costs of this investment (and also probably the additional electricity costs for running the node).

Maybe it’s not the right approach but, on top of this, I also feel rewarded by embarking in a project that seems to me more reasonable from an environmental standpoint than the constant building of new datacenters, and which also strives to safeguard the ownership and privacy of data. This may sound philosophical but these aspects are nowadays very appealing.

Finally and to connect with the current thread (I do not have a huge IT background…) I trust that you will find a well-enough balance to make your solution attractive enough to embark people like me in your journey for the long term. In any case it would be good to be able to easily monitor whether or not we are meeting the downtime requirement and, if close to the limit, what is preventing from meeting it by a larger margin.

1 Like

Just to add my $0.02,
I am making a profit even after taking into account cost for replacing drives and energy used. I’m using very old domestic grade equipment which has “depreciated” to the extent it’s worthless to sell even on ebay ( I’d be lucky to get enought to cover P&P)

My Example:
My “Storj server” is an old AMD Athlon x4 840 with a 4Tb hdd, full power-draw from the wall is around 40 watts. So it takes an average of 5 months income from Storj to cover a years worth of power usage. The I’ve set aside another 4 months of income specificaly for Hardware replacements/upgrades.

So I have 3 months of “profit” per year. ( NOTE This is NOT including the “surge” pricing which skews some peoples income expectations, that in itself was a really welcomes bonus which, in my case, went on a new PC case for my main rig.)

Even if I included a maintenance charge for my input which accounts to around 20 mins per month for reboots and log checking I’m still “in profit”

If the node is set up properly and securely the project is , in my opinion, a "set and forget "project.

The “profit” in storj is to be able to reuse old outdated equipment you currently have to generate a modest income. Unlike most project it has never been sold as a way to “generate a livable income” (in fact Storj is at pains to say the contrary every time a post about income is made.)

Yes their are people renting out servers to host a node or dozen, thinking they will “get rich quick” but quickly find out that is far from the case. Their are a few with access to “spare” resources in data centers which they use to run a nodes but since they are using “spare” resources not fully under their control.

Well that’s my input.


I have invested about £600 in hardware for storagenodes and so far got $5 return (in three months) but I’ve had about £1000 of fun!

My 5 nodes use about 25W so electricity will be ~£50 so I might make £10 this year or up to £300 if the nodes fill up.

1 Like

When I started with STORJ I had an odroid hc2 as Homserver which was bothering me all the time because it was too slow, not powerful enough, etc…
So STORJ gave me a good excuse to buy a new system for ~300$ that runs on a Ryzen 2600 with 16GB DDR4 and zfs raid. However I am using old drives I had lying around otherwise this would be a bigger investment.
I started early with STORJ so that was definitely an advantage and so far the payments easily cover the investment costs + electricity. So I really can’t complain. Especially because I am using the homeserver for a lot of things anyway.
I don’t count internet costs because I’d pay for that anyway.
Time used is just fun. Never had any problems with my nodes so even without monitoring they are basically “run and forget”.


The whole context of this was in the downtime dq discussion. People were talking about what could be done to guarantee SNO up time. The point I was making is most people are not breaking even with a bare minimum setup. We can sit here and debate what costs should or should not be counted. I try to take a logical look that all costs count if you really want to look at it from a business standpoint. I am not looking at if from a standpoint of this is just a fun project to be a part of. So adding extra things like backup ISP, redundant hardware, ect just makes no sense for what the payout would be. We will see what the downtown dq looks like in the end. I think people will look at this a little differently if your node gets dq for a couple unfortunate downtime events and you loose your held back amount.

I say this with having 99.075% over the past 30 days that includes downtime for ISP being unavailable, nodes did not start correctly after patching, ect. I also look at what could happen. The ISP could be down for several days (no sla on consumer connections), could have hardware failure were it takes a week to get a part delivered from ebay ect. To many scenarios to cover.

That is why I also asked if any type of modeling has been done or will be done with what is currently being proposed to know how may nodes would have been dq over the past thirty days if the rules where in place. Should probably go further back then that do the rules of hold back. Like i said I think if you loose your held back amount after 12 months do to a couple unfortunate things that cause downtime you will look at this a little differently.

Of course DQ will make you look at things differently. But as of now we can’t tell how much really has to happen until you get DQed. STORJ is aware of these issues and if they choose their numbers right, the suspension mode could be enough to prevent unfortunate DQs. So in the typical scenario you should only get DQed if your node lost data due to hardware failure. (or if someone tries to cheat the system for whatever reason…). At least that is the goal.

It also looks like that being offline at the wrong moment may get counted as “lost data”.

It’s seems that STORJ don’t have a trafic due to missing clients, don’t publish any statistics and say that don’t have any statistic but in the same time they say that will support SNO by test traffic for posibillity to make payments. Which level of rewards for SNO will be supported like minimal ? I have some expanses for support my nodes and after several monthes I start think that my expanses stable less then income payments - it’s not a business !!! Adittionaly I think that if rule “not more then 5 hours per month” will be applied then not one SNO will not received own holded money - during the year we all can have this cases due to missing internet channel or electricity or somthing else…

@Pentium100 @vargrant let’s not speculate about things that aren’t announced yet.

It’s a blueprint under discussion and review. There’s no final decision yet so it doesn’t help building opinions on what “it looks like”.

Nobody says that 5 hours per month will get you DQed. In fact the new blueprint will allow a lot more.
We’ll see how much more once they have decided how the new system looks.
Until then there is no point in making a big fuzz over something that is not enforced and not even decided.