StorJ price increase Warning

STORJ was set up to fail from the beginning. The economics never made sense.

First, there is the STORJ token. This is basically just an ICO, or runaway money.
We can completely ignore that.

So what is the unique selling point of STORJ?
It only uses unused hardware.

That is it. And a great idea, if you ask me. Instead of building large datacenters like Backblaze, you simply pay nodes to do the work for you. In theory and on paper, that comes with multiple advantages. You get awesome speed, and awesome peering, all around the world. You don’t have to pay for hardware. You just pay a small amount to node operators. And because these node operators only use unused hardware, you can pay them a very small amount. Because for them it is either no money at all and just having empty space vs. free money.

In reality, there are multiple issues.

Software quality:
To distribute these chunks is not an easy task apparently. STORJ software quality was always extremely poor. I am not a programmer by any means, but even I knew that sync writing data to my node is absolutely stupid and makes no sense whatsoever. I even told them that in the forum. It took them over two years to change to async writes. Nobody ever really noticed how poor the software was, because there were no customers. When the first one arrived (that was before their 30-day minimum, a customer wanted to store CCTV for two weeks), the non-ZFS nodes pretty much all crashed. To be fair, I don’t know the current situation. Since there is still not really data in the STORJ network, nobody can know.

S3 or native:
This is by far the biggest issue that STORJ faces. They offer a native integration for Linux. But who wants to run that, when you can just use S3? Turns out, almost nobody. The funniest thing is, that even the posterboy partner TrueNAS could not be arsed to implement it :slight_smile: Naaahhh, we will just use S3. But using S3 comes with a whole lot of issues. The decentralization is gone. You are now using a STORJ satellite, that uses the nodes as a storage backend. The satellite is not decentralized, has to pay expensive traffic, is slow, is expensive, the list goes on.

So with that knowledge, let us look at the costs for STORJ.
You pay nodes 2$ per TB. Since you don’t have much escrow and a node could leave at any minute, you distribute data by the factor 2.2. So you pay nodes 4.4$ per TB. You have to run the satellites that keep track of where data is stored. You also have to run the satellites to get storage from the nodes, to the customers.
We don’t know the exact costs there.

So per TB you have the costs of 4.4$ + satellite.
Now you have to make a package for customers. You decide on 7$ per TB stored and also for traffic. Do you think that is enough to break even? I highly doubt it. I think you are still running on VC money.

The last and biggest issue; who is STORJ for?
Nobody seems to have an answer for that. And that is why there are almost no paying customers and no growth. 60PB and that is including their SOC2 datacenters. I would not even trust that small number, since they did depatable stuff there too in the past (free user data, test data).

According to the forums, storj is is faster than Amazon S3, has free geodistribution, included CDN, so at 7$ it is basically a steal. The problem with that claim is, that as soon as you start using S3, this flies out the windows (maybe even with native, I have not tested it).

homelabber: Want a cheap S3 backup for his TrueNAS. Backblaze is cheaper, more trustworthy.

S3 VMs/VPS/Apps: Too slow, since it is not “local” and needs peering.

CDN: To expensive at 7$ per TB. Compare that with Cloudflare, it is not even close. TrueNAS uses it for the ISOs. Probably got a good deal on that. And the download performance for me personally tanked since they switched to STORJ. I am getting 20MB/s right know, and that is a good day.

To this day, I have not heard any convincing argument, why anyone should use STORJ and not some better alternative!