I’m willing to adopt orphan nodes! They’re worth more than just their graceful-exit tokens 
I’m reading storj white paper for the first time:
Our objective is to aggressively compete in the wider cloud storage industry and bring
decentralized cloud storage into the mainstream. Until a decentralized cloud storage
protocol becomes widely adopted, Amazon S3 compatibility creates a graceful transition path from centralized providers by alleviating many switching costs for our users...
You can borrow Microsoft 3E strategy, but in a good way, write a library support multiple protocol (S3/storj/a few more..) in top 10 hot languages, promote it, have a foot hold in community, the switch will be painless, random idea…
It sounds more like a mistake. I will forward that feedback to the team and hopefully they can rework the message.
Also thanks to @alpharabbit for pointing it out first. I would bet this is just a copy paste mistake from the object mount tab. For object mount we don’t have a self serve service and contact sales is the only solution for now. It looks like they just updated the text and used the same layout including the contact sales button. Small mistakes like that can happen. Thank you for pointing it out. We should be able to correct that within a few days.
Then I apologize for overreaction!
I fully agree that is how it should get implemented one day. Instead of sending the storage node payout there should be a step that checks if any of the payout addresses is matching a customer deposit address and if so remove it from the payout list and instead add credits to the customer account directly without any transaction on the blockchain.
That luxury solution is still in the backlog. We just didn’t had the time to implement it. Meanwhile the initial cheap workaround remains active for way longer than anyone expected.
So where does a free user download object mount? Your site still requires contacting a staff member to download the software lol. i’ve agreed to updated pricing, you’ve told me I get 2 free licenses, but there’s no way to access and use object mount.
As someone that has supported Storj and wants to continue, you guys make this so difficult to do so. It’s wild.
We have a dedicated Gateway-ST self-sufficient binary, why duplicate this in uplink?
These licenses and download links should become available after July 1, 2026, when a new pricing become in effect, then there would be an instruction. But basically:
- Navigate to Object Mount in your Storj Console
- Create S3 credentials
- Download Object Mount and follow the guide Windows Native Installation Guide - Storj Docs or Installation Guide for Native macOS App - Storj Docs
- When it will be installed, it will automatically mount either a selected bucket(s) or all buckets on a one disk in your OS.
While you have more than 0 paid invoices it will work.
If provided S3 credentials would be not Storj, it will demand to install a license and mount nothing.
Update:
Please take a look:
Ok so wait a month, thank you.
This isn’t using native integration?
Out of the box it is using S3 credentials. There have been some experiments with native integration. I didn’t follow the progress on that.
@jammerdan above linked my post where I drafted a possible scenario of this type. There seemed to be more criticism than supporters 
Though, please still upvote it if it would be something you’d actually want to use (not: “maybe consider”).
The number of /24 blocks remains relatively stable, so it’s likely same node operators spinning up new nodes.
This strategy has been tested by Sun Microsystems. It… didn’t work.
That’s interesting: it looks like May 2025 had 11895 /24s, and May 2026 had 12844. So we gained around 1000 subnets over the same year we added around 4500 nodes.
So yeah there’s probably lots of new-nodes for old-SNOs… but that’s still a lot of fresh /24s as well.
However it looks around 70% of that subnet gain was within 88.208.243.0 (most nodes-in-subnet)… which is in England? The UK was also #5 in node-count in those May 2025 numbers… but have moved up to #4 in May 2026.
Some SNO in England is building a monster setup???
This is select network (IONOS). 
root@server030:~# curl -s http://th3van.dk | sed -n '27p'
Number of backup nodes online @ backup-parking-slot001.storj.dk : 1189/1451
root@server030:~# ping backup-parking-slot001.storj.dk -c 3
PING backup-parking-slot001.storj.dk (88.208.243.158) 56(84) bytes of data.
64 bytes from backup001.storj.dk (88.208.243.158): icmp_seq=1 ttl=57 time=32.7 ms
64 bytes from backup001.storj.dk (88.208.243.158): icmp_seq=2 ttl=57 time=32.5 ms
64 bytes from backup001.storj.dk (88.208.243.158): icmp_seq=3 ttl=57 time=32.6 ms
--- backup-parking-slot001.storj.dk ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 32.524/32.589/32.666/0.058 ms
root@server030:~#
root@storj-1k-nodes:/# ps -awx | grep -v "grep" | grep -c "/disk001/storj/storagenode/storagenode"
1000
root@storj-1k-nodes:/# free -h
total used free shared buff/cache available
Mem: 62Gi 30Gi 1.4Gi 844Ki 31Gi 32Gi
Swap: 8.0Gi 7.9Gi 150Mi
root@storj-1k-nodes:/# df -h /disk001 # (nvme)
Filesystem Size Used Avail Use% Mounted on
/dev/sda3 2.8T 2.3T 471G 84% /disk001
root@storj-1k-nodes:/# hwinfo --short --cpu | awk 'NR==2' | sed 's/^ *//'
AMD Ryzen Threadripper 3970X 32-Core Processor, 3693 MHz (12 x vCPU)
root@storj-1k-nodes:/# uptime
14:38:04 up 80 days, 3:41, 1 user, load average: 0.43, 0.29, 0.21
Th3Van.dk
It has a read-only native integration
The mystery is solved, thank you!
But why on earth do you have over 1000 pre-vetted/post-withholding nodes idling? Aren’t they using hundreds of GB of RAM just being turned on?
Or maybe your company considered being part of Select… so you have them all ready to transplant to spare HDDs?
I really, really look up to Th3Van. Somethings we do differently, but over the time, we do much of the same things.
I also have ~200 backup nodes ready, online and waiting. Homie upgrade to 3TB of memory back in '24, and he’s all memtables. RAM allocation is not really a concern for him. As far as I remember, all backup nodes are on a seperate system running SSD only, on a single /24 .
I created all my backup nodes while refining my “create multiple node identities in one go” script. Whenever I get access to a new IP address, I run 5-10 nodes at that location, and pluck them from the backup node pool
I’m not constantly filling disks, or getting new IPs, but I understand the idea.
I had heard of people making lots of new nodes… back when the pool of pre-vetted upload capacity was more generous. But if every node is around ~175-200MB of RAM I don’t have that to spare. Even 200 backup nodes sounds crazy! 
What a tone-deaf reply – essentially, “if you don’t like it, leave – tough.” 
It’s sad to see that such a great service is throwing its “small” customers summarily out the door. I was paying around $10/mo for under 3TB of personal/family data and now have to migrate all that data elsewhere (which will take days or more over my ISP - direct cloud transfers far too expensive). While business costs increasing and being passed onto customers of course makes sense, such a drastic change practically overnight does not. This really makes Storj management look very bad to this user. I went with Storj last year over B2 and others because of the slightly cheaper cost and the distributed nature – what a shame to be thrown under the bus like this. Existing “small” customers should have been exempted from this $50 minimum.