Storage Nodes Full of Test Data

This does not apply to all smaller nodes. I myself have a small node running for much more than 3 months now that is not even half full yet. Not everyone running nodes has great bandwidth and RAM.

Sure the node will be disqualified, but when I have a few MB/day instead of tens of GB/days of egress trafic when I have some space available, the count is quickly made. Shut down this unprofitable node and apply for a new one.
The problem is the business model used by Storj, they build a completely new technology and want to align their price with AWS. They should rather based the pricing on the service cost. Bandwidth cost for the nodes is the same if they checkthere mail once a day or upload 130 TB/month (400Mbs * 1 month). But storage with a 97% availability has a cost, electricity, HDD, computer. The egress should be drastically reduce, a $1/TB would be sufficient and the storage greatly increased like $15/TB/month this would make a service using the unused bandwidth, rather than been a system specialized in cold storage. With a serious problem in garbage collection.

3 Likes

This is a real issue for me. Once my 3TB node filled up, I am getting no egress, thus no incentive to run the node anymore. The amount received for the storage space is too little.

The escrow is going to keep SNOs interested for so long, eventually they are going to hard exit.

The combination of escrow and lack of graceful exit is creating a business model that is not going to be relevant for long.

3 Likes

Yes absolutely, currently the most profitable option is to close your node when its full and open a new one.

2 Likes

Dear @lex-1 and @fikros,

please do not forget that Storj informed us, as Storage Node Operators, to run their software on computers which are already running 24/7. They explicitly asked us not to buy any new hardware to run nodes, but to run it on already present hardware which is not used to its 100%. They are full aware that it is not worth it in any other case than that.

In my opinion Storj is not using money to motivate us, I believe they are looking for SNOs which are passionate by the idea of creating an alternative to Google, Microsoft and Amazon. Or people enthusiastic of a new technology.

Having said that I do agree on the fact that having a node which shows activity, any kind of activity, is much more motivating than anything else, as it shows that the project is working. I also have a node which I am not sure if it is doing any good or not, I also would like to see some real ingress/egress data. I am checking everyday my router logs hoping to see the bandwidth use skyrocket (this will cause another problem then!) or the check if my infrastructure can cope with Storj requirements. But we are not there yet.

I believe we all are on the same boat and have to wait until the projects starts to pick some traction from customers all over the world. Basically we have to trust the company to make a good commercial strategy so to attract many customers and quickly. Still, since we are not in production yet, we need some more patience.

3 Likes

You may want to read this No downloads on full node and also note that graceful exit is close to being released. We are sorry but you cannot expect everything to be fully implemented when we are still in beta.

Guys, don’t get me (and most certainly @lex-1 ) wrong, we are here as alpha/beta and early adopters. Our goal is to give feedback, where feedback is needed.

Storj is a business, and is competing with other businesses, like amazon, google, azure/microsoft, dropbox and others.

So, in order to become a veritable competitor in this market it should make sense business-wise.

Storj needs to make sure the data exists on the network, accounting redundancy, at all times. It also needs to make sure the data is available for download or upload in a certain time-frame, therefore it needs the bandwidth.

This means that it must pay SNOs for these services.

Don’t imagine, for one second, that SNOs would do this for passion’s sake. Most SNOs do it for the money, hell, they are even trying to abuse the system for the money. I’m sure that explicit examples of how the system (Both V2 and V3) can be abused aren’t needed, and are known to the dev team.

No one is going to pay Storj for storage space knowing their business model is based on passion. If I, as a business, am going to consider Storj for my storage needs, I need to make sure it is based on a sound principle, not on a hit-and-run model.

So, my humble opinion is, if you want to test the business model, either pay SNOs upfront (no escrow), or have a working graceful exit / scale-down solution BEFORE release. This way the storage price is being determined by the SNOs needs and willingness to provide, not on a model that is based on ransom, MLM, or other shady principles.

Let me reiterate, I have no stake in this, I can and will delete my node at will, but am genuinely interested in Storj/Tardigrade as as service.

4 Likes

@fikros I fully agree with you.

Dear @fikros,

thanks a lot for your message: it definitely makes me think and most probably I am one of the few that has little interest in the money. The computer on which Storj is running is already busy with the folding at home project. Basically the only thing which was not used by F@H was the HDD and the internet connection. This is why I started looking for an HDD sharing project and this is how I came across V2 years ago. So, basically, since I am already volounteering for another project, most probably I am already coming from a biased background. :smiley:

If I was considering the money an important topic I would not invest that much effort in a project which is paying in Storj, an unknown cryptocurrency, which, I believe, will be very expensive to exchange in a useful currency. Add to that the volatility of cryptocurrencies… waaay too risky. Am I wrong?

I mean, the only thing I could use storj cryptocurrency for, is to buy some storage service, is it not? Otherwise if I wanted to use that money to pay my electricity bill I should exchange Storj to Ethereum and once more to €/$… I have no knowledge on how it works, but I always had the idea that all this steps will be super expensive.

Let me add some comments to your message.

They are, correct? What you suggest is that they should pay more, if I understand you right? On the other hand I remember that when they were looking for further SNOs they did increase the pay (following the Uber model, is it not?).

You are right. But money alone is not enough to motivate people either and if not managed carefully could istigate some hit and run behaviour as well. And once more: if you have your servers already running 24/7 what is the expense that you have to cover? You are already covering the costs for electricity, already paying internet connection, etc.

I find more precious the time that we are investing in controlling the node and providing feedbacks. This is really precious in my opinion. And for that we are not getting any reward in terms of money… Not to mention the time I have invested in moving from docker (on windows), to docker on linux and after that back to Windows GUI. Jeez, such a waste of time (me being a noob did not help either! ahah).
The only reward we may have versus future SNOs is that, being here “since the beginning” will give us some advantage in terms of traffic (inbound/egress). I do not see why SNOs would invest so much time in checking the node, troubleshooting, writing feedbacks if not by being motivated of being part of a new project.

So, in my opinion, it is the beta testing that we are doing which is the big investment that we have decided (more or less consciously) to do and apparently storj successfully managed to get our trust on that. I have to admit though, that I have already thought more than once about quitting, as sometimes, getting the node to work properly and learning/testing new things ended up in being really time consuming and, when the offline time limit was active, quite stressful (e.g. doing all the migrations back and forth while trying to be offline for the minimum amount of time). Thankfully @Alexey was always there and very patient. :slight_smile:

In conclusion, the best advice that comes to my mind, is to suggest everyone to join our team on F@H and start volunteering for real! :rofl: Visit Swiss Folding Team for details.

2 Likes

Thanks for your supportive comments. Just fyi, it is not excessively expensive to exchange STORJ for other crypto such as Bitcoin or Ethereum, and go from there to your local fiat currency. You can do it easily by following these instructions. I also want to commend you for participating in FAH. I have been doing the same volunteering for foldingcoin.net myself (yes, you can help cure diseases and earn crypto by participating in Folding@Home). But that is off-topic for this forum. Keep up the good work!

3 Likes

@aseegy I see your points and if Storj was developed to build a free file storage system, like Linux is a free OS I would agree with you but Storj is a commercial enterprise with the goal to compete with the big data-centers and make big profits. My suggestion was toward making it more profitable for Storj and the SNO’s.
You are never wasting your time when learning, only when maintaining a commercial system for free.
So far been an early adopter only fill-up your disk and block all trafic.
The tokens will be worthless if the project fail but if Storj is successful its gold.

2 Likes

@fikros I do agree with your view.

Right now, even though I joined the beta out of curiosity with a spare 1TB disk I already had, I would consider carrying on and being part of the story on a long term if what StorjLab pay is an interesting enough incentive.
The project is really interesting technically but I also like the idea of competing with centralized solution giants like Amazon, Google… In a way, that’s brilliant and encouraging to be part of that as an SNO :slight_smile:
But passion alone won’t drive me further than just trying this out for a few months if it does not get me a bit of money along the way, I’ll be honest.

In my opinion, (and if I’m not mistaking) considering that bandwidth is sold around 45$/TB to users, it feels kind of fair to pay SNOs approximately half of that (20$/TB).

On the other hand, storage space is sold around 10$/TB to users, but SNOs get only a fraction of that (1.5$/TB), which is far from being enough for considering maintaining a fair amount of storage space. In my humble opinion.

We received an e-mail from StorjLab recently claiming that they are going to need way more space for future users, considering the number of developers on the waitlist, encouraging us to add more storage and invite our friends.
I think one fair way to motivate people for adding more space is to make this profitable:
What about paying SNOs at the very least 5~6$ per stored TB?
Especially if it becomes kind of the only revenue when a node is at its maximum storage capacity.

Anyways, great project, brilliant technology I’m sure :slightly_smiling_face: :+1:
Just need a way to make sure SNOs want to stay.

3 Likes

hi @Pac
thanks for sharing your thoughts so eloquently – we love to hear from all viewpoints! The pricing model is something that we as a company spent a lot of time on. Since the flow of money is the lifeblood of a business we tried to look at it from as many angles as possible.
I understand and agree with you that each storage node operator will need to decide for themselves whether the arrangement is suitable to their needs. What we try to do is offer as much transparency as possible around what that arrangement consists of, so that people can be equipped to make that decision according to thier individuals prioirites and logic.

I do hope that you stay on!

I can also pass your comments on internally. Although Im not the person who runs the pricing model, you can engage with them in the live section of our Q&A section of the next town hall and hear a response form the horse’s mouth. The way to do that is to email ask@storj.io so your feedback can be imported to the list of featured comments.

Cheers for a great approach! Keep up the good work! Please keep the pricing model as a priority, because it’s a make-or-break matter.

1 Like

Hey @Pac

I think you’re overlooking something essential here. Every 1TB a user stores on the network needs to be redundantly stored with an expansion rate of roughly 2.7x. So that means you need to divide that $10 by 2.7 to know how much Storj roughly receives for each 1TB stored on nodes. So roughly $3.70 per TB. I think you’ll agree that similar to bandwidth a little below half is pretty fair or at least much more fair than your initial calculation.

A part of the income is also shared with open source partners, so Stroj is not pocketing the rest of it entirely either.

2 Likes

Hello @BrightSilence, thank you for this clear explanation, I did overlook that.
As a new user, I’m still getting up to speed with regards to all the STORJ ecosystem.

This said, maybe there are a few things that can be adjusted and/or discussed with the community. For instance:

  • Storj do not pay us in dollars/euros, but in STORJ tokens that they mine themselves (I guess?). Doesn’t that mean that SNOs’ payments could be more or less unrelated to clients’ payments?
  • Storj are always going to be paid via trafic, globally (coming from all clients), whereas it might not be the case for some SNOs if their disks are full. Maybe most of the money collected for storing data could be paid to SNOs, like 3$ per stored TB.

Just trying to exchange thoughts :slight_smile:

2 posts were split to a new topic: Incentives and payments for making SNOs stay

STORJ doesn’t mine STORJ tokens. There is a fixed total amount that can never change. They have large reserves for now but when that’s gone they will be forced to buy them back themselves to pay SNOs. You should see the token as just an intermediate, it’s not a magic way for them to create money where there is none.

This could be an option, though there is something to say for paying nodes for the actual thing that brings value. I wouldn’t worry about the egress problem for nodes that are full too much as this is clearly an artifact of how specific test patterns are handled and this most likely will not be the case when most data on the network is actual customer data. So to a certain extent I think you’re trying to fix a problem that will no longer exist when the network is in full use.

2 Likes

@BrightSilence Thanks for these insights. Maybe you’re right and having disks full of data won’t be a problem, let’s hope so :slight_smile:

1 Like

9 posts were split to a new topic: Tokens flow and usage