Disclaimer: This reflects my own point of view. This is not an investment advice. Do your own research. You are free and encouraged to share your (opposite) opinion with me in comments section if you are able to discuss in a peaceful and respective manner :)
I usually spend a lot of time discovering, learning and trying new technologies (some are new for me, others are new for everybody). That is why I spend a lot of time on Blockchain topics. This is not (only) because of financial incentives but mostly because this new technology is a breeding ground for absolutely exciting projects and disruptive approaches. Blockchain is to the 2010s & 2020s what the World Wide Web is to the 90s. Do you realize all the improvements, innovation and new mindset that the WWW brought since its beginning? We are just at the beginning of what the Blockchain will bring, for better and for worse.
One day several years ago, I was thinking about what the use cases the blockchain could address. Thanks to its decentralized attribute, I was wondering "could it be used for backup purposes?". A file could be split into parts and, as any block in the Bitcoin or Ethereum network, these parts would be stored on multiple peers on the decentralized network. It would be almost impossible for a file to be lost. By looking for something like this, I discovered the fascinating world of decentralized storage projects: first Storj and Skynet (formerly Sia), then IPFS, Filecoin. There are more. Though, they don't really work how I described it earlier, they are even better.
That's how I started with Sia and Storj.
I've been using Storj since August 2019. And I definitely adopted it against Sia.
After several discussions online, the same questions came: is it worth it? Why not Sia? Is it secure?
This article is not about comparing both, neither comparing Storj with other techs. It is just about Storj. Maybe a forthcoming article will deep into the similarities and differences between both. This article aims at answering some fundamental questions before investing time and money in Storj project as an active Storage Node Operator (SNO). So if you are hesitating between both or you are just curious about Storj, it will give you some insight about how much you can earn and what it takes to get involved in it.
What is Storj? What need does it meet?
The underlying and fundamental question is: Is the market on which Storj operates profitable?
Let's look at it.
Basically, Storj and other competitors like Filecoin or Sia are each (more or less) a decentralized network that allows to store data with a high degree of durability and security. I like to describe them as the "Uber" of the Cloud Storage Market.
Each of these services has its own characteristics (and and its own architecture with specific technologies) but they all basically answer the same needs:
- Privacy of the data
- (Very) High-Availability
- (Very) High Durability
- Cheap storage price
Of course, not all Decentralized Storage Providers (DSP) have the same level of maturity on each of these characteristics.
Major Cloud Providers already provide some level of these benefits. But they have a drawback that they will never be able to fix: they are all centered around the same company, the same actor, even if their architecture is highly distributed and even if they are very good at it. These centralized providers have centralized configurations (even if not all), may be under legal restrictions or may have some individual interests that are against customers.
By using Decentralized Storage Providers/networks, customers don't have to trust a single actor to take care about their data. Their data is stored on multiple actors storage, that don't know or trust each other. As a consequence, the security risk is much lower because a hacker would need to hack several nodes, across several (and probably random) regions.
When a newcomer disrupts an existing market, like Uber did, it is because of an unmet economic need, or a not well met need. So what is it exactly in storage field? See by yourself:
- Leaky AWS S3 bucket once again at centre of data breach
- Microsoft Azure cloud vulnerability is the ‘worst you can imagine
- Millions of websites offline after fire at French cloud services firm
There is a market opportunity because of lack of privacy, lack of confidence in availability and durability.
The answer is: there is a market opportunity because of lack of privacy, lack of confidence in availability and durability. Oh and guess what, cloud storage is actually not cheap at all compared to what DSP can provide.
Though, even if there is a huge opportunity and a real need, what I can see today is that this market is very small at the moment. It is very difficult to have figures about it. I don't know any big company that already pushed a significant amount of its production data into these DSP. In my opinion, there are 3 main reasons of this, that should not scare you:
1. Accountability: Big companies need someone to turn to if something goes wrong. With 100% decentralized services, it is impossible (as it is impossible to ask for "After-sale service" if you lose your private key to get your blockchain wallet).
2. Technology maturity: Even if I think that Storj is very good at it (it didn't lose any single data since it launched), the decentralized storage is something too new to be widely adopted right now by companies. Good technical arguments will help but these companies need also time to get involved in it.
3. Change in paradigm: Decentralized storage may change how these companies deal with data. They have processes, habits (internal and external) and regulation that are real obstacles for a wide adoption.
Existing (centralized) cloud storage is about to be disrupted just because the future will be decentralized.
In a few words: existing (centralized) cloud storage market is not suiting the real needs of both small and big companies. Existing cloud storage is about to be disrupted just because the future will be decentralized. Then, Decentralized Storage Cloud is a huge opportunity, even if it has some big rational and irrational challenges to face.
How does it work under the hood?
The purpose of this article is not to get into the technical details of how Storj works.
But before deciding to be a part of the network and to invest some money in it, you need to understand how it works at a high level.
As you can see on the following high-level diagram from Storj whitepaper, Storj is a set of 3 main actors.
- The Customer application represents the customers themselves, those who are paying for their data to be stored on the network.
- The Storage Nodes are the one who actually store the data, as the Uber drivers drive customers. To be more accurate, they store all together pieces of the initial file. People who run Storage Nodes are called Storage Node Operators (SNO).
- The Satellites currently represent the underlying infrastructure of the market. Without it, the market can't exist, as without Uber platform, Uber drivers and customers can't meet each other. This part is currently operated by Storj Labs (but it may change in a more or less near future).
- Above those 3 actors, there is what I call the Economics layer or Market layer. This is the way the network makes it possible for storage providers and customers to meet.
Breaking News: the first 3 parts are not powered by the Blockchain. Storj is mostly NOT a blockchain project! Other articles that cliam the opposite just didn't understand how it works. Don't take my word for it, read the Whitepaper, ask the community. Though, Ethereum blockchain is used for the Market layer in order to pay for the service in ERC-20 STORJ token (and to get paid in it).
Storj network is not powered by the Blockchain. But the Blockchain brings the Market layer on top of the network in a decentralized way, which is critical for a Decentralized Storage Provider.
The overall process of uploading data is very secure:
1. Data is encrypted locally on the customer side
2. Data is split into 80 pieces
3. Data is sent and distributed randomly at a global scale
When the customer needs to get back his data, only 29 pieces are needed. The network itself performs automatic repair process if too many nodes go offline.
You may have noticed something very important: yes, Storj is not fully decentralized. At least fort the moment.
Indeed, Satellites are essential to connect SNO and Customers. Satellites are redundant but should they become unavailable, your data is still safe (satellites don't hold encryption). Once up again, you can get back access to your data.
This characteristic is fully assumed by Storj Labs.
As they explained it in their blog, a Decentralized Storage Cloud has to choose between being Economically sustainable, Enterprise Grade and 100% Decentralized.
This is their strategy. They think that they need to be first economically sustainable and provide Enterprise grade service before being fully decentralized. That's what make them so unique in this market.
It is crucial for the network to be economically sustainable (for Storj Labs and for SNO). In other words: if network contributors (SNO for Storj) are not well paid, they will leave the network, making it less decentralized or disappear. And that is one of the reasons why I choose not to continue with Sia.
If you need some reassurance regarding the Satellites, you can find here some details on Storj blog.
Storj first aims to be economically sustainable and provide Enterprise grade service before being fully decentralized. In a sense, this compromise strengthens the network because more SNO are incentivized to be part of it.
Comparison with other Decentralized Cloud Providers
I'won't lie to you: It is very difficult to have a detailed and objective overview of the whole Decentralized Cloud Storage market. Even if I tried Sia at the beginning and got into the technical and economical details about other solutions, I have much more experience on Storj than other Decentralized Cloud Storage projects. But if you want to compare, you have to consider several parameters:
1. Technical robustness: it is obvious. Of course, not everybody can assess it easily. Underlying questions may be: How decentralized is it? Is there any breach/major incident precedent? Is it well documented?
2. Vision & Roadmap visibility: What "dream" the project has? What are the project priorities? What goal the team tries to reach? What are the main improvements in the roadmap? Do you feel that the next improvements will bring value to the customers?
3. Community: Is the community active? Helpful? How does the development team involve the community (level of communication, project and financial transparency, decision-making/main orientation process, etc.)?
4. Economics: attractive incentives are very important because you don't want to get involved in it if it doesn't pay off. In the end, nobody will and the quality of service will be at risk.
5. Development strategy: What is the development team strategy regarding the trilemma "Economically sustainable, Enterprise Grade and 100% Decentralized."? Are you comfortable with the compromise the project team is doing? Does the roadmap plan to change it? How?
If you are asking me, my opinion is that Storj is better than Skynet/Sia because it gives attractive incentives to the individual suppliers. By doing this, it makes its network infrastructure more resilient, even if it is less decentralized than Skynet.
I would also consider it better than Filecoin because it is easier to get into it as a Storage provider. High requirements on Filecoin may make it less decentralized (see the Blockchain trilemma).
Storj is better than Skynet/Sia because it gives attractive incentives to the individual suppliers. By doing this, it makes its network infrastructure more resilient, even if it is less decentralized than Skynet.
So is it worth it?
Now you have a quite good overview of the market opportunities, you may ask: is it worth it in the end? How much will I earn?
Let's take my own experience with my main disk, which has been running for 2 years.
- 1 HDD disk (Western Digital Red Desktop 3 TB): $ 126.98
- Power: between monthly $0.3 and $0.6 based on ~5W consumption
- Total net earnings (net earnings paid by Storj - investment costs - operational costs): $369.15
- Total ROI: 290.70%
- Annualized ROI: 166.71%
If you want the details, here is my Earnings Storj dashboard:
Some remarks and advice:
- Of course, you will need a system to run the node. I already had an HP Microserver Gen8 with enough space so I didn't take this into account. Though, you could build a similar system with a dedicated server, small computer (such as a NUC) or even a Raspberry Pi with appropriate power supply.
- Part of your earnings will be put aside as a "held amount" during the first months. This acts as a "stake" although you don't have to bring it before joining the network, which is very comfortable! More details here.
- As an SNO, as I explained in an previous article, you should be ready to invest your own time in order to maintain the system. Sometimes, it could be a non negligible amount of time.
Do NOT go with SMR and SSD disks. Even if they are cheaper, they deliver poor I/O performance (that could greatly impact your earnings) and their lifetime is very low compared to NAS/Enterprise grade internal HDD. The best ones are probably SCSi disks but they are more expansive. I hope the detailed financial stats I gave you above will help you to do your own business case!
Is Storj a project you should involve in as an SNO?
All depends on your profile and what you are looking for. If you are looking for an "Invest and forget" strategy, forget it.
If you are a little bit techy and want to try a very promising project, give it a try! If you follow some good practices, it will be profitable.
- 2021-09-01: fix minor typos