Does anyone know what is the size of an average Storj Node?

Node Operators
21

15.7 TB on my 20 month old node, have had some extended downtime along the way, so could have had a bit more, but have had various issues along the way.

it’s never been out of capacity and usually had successrates in the top end of 90% running on a 1gbit sync fiber.

good successrates are in my experience down to low latency in general, and low latency means top tier ingress, some disks aren’t very good for storj like SMR tho not always the case.

generally latency is something like a 50 / 50 split between disk caused latency and internet latency, ofc this depends on many factors and thus isn’t easy to pin down and both will vary greatly depending on disk workload latency to different regions.

from testing early on when i started we pretty much determined that ingress is the same across all optimal functioning nodes, no special hardware required.
infact it seems a lot like one of the issues for setups with poor performance is due to the hardware not being able to keep up, usually disk io related as storage nodes is very iops heavy.

internet on the other hand seems very low priority so long as one isn’t running like less than the recommended 25/5 Mbit connection, my storagenode usually pulls like 200kbyte ingress and
…
image
less than 800kb/s egress avg over the last 4-5 months

so thats less than 10Mbit Egress and 2Mbit Ingress, so well below the recommended minimum.
ofc it can be easy to load a 25mbit download connection, but i would also argue that these are quite few and far between for SNO’s.

have a few newer nodes that are still vetting, and been vetting for like a year or so… only a couple TB on those.

oh yeah and my 20 month node has been around the same size for ages now… most likely due to the occasional downtime over 4 hours.

so long storjy short, keep low latency and check your successrates and you should get 500gb ish ingress on a vetted node pr month… ofc takes forever to vet these days… so new SNO’s or nodes certainly has a big handicap.

and the more data you get the more the deletes will cancel out your ingress.
much of my data was gained during the first 3-6 months due to the final storj network stress test data.

1 Like
22

I use a Synology DS920+.

23

I run it as a cool project. Want to support it. I actually have 4x16TB in SHR1. Just storing movies and photos. Got it right before the chia-inflated prices.

24

Me.

NAME   USED  AVAIL     REFER  MOUNTPOINT
stor  33.2T  1.94T      240K  /stor

This is the server where the Storj node VM runs. The rest of the space is taken by chia (kind-of pointless now) and backups of some of my devices.

This server has 12 hard drives (6x4TB, 6x6TB), arranged in two raidz2 vdevs. I started with the 4TB drives, then, when the node filled them up, I bought additional drives and added them to the pool

My file server (the one where I keep my movies and backups of my other devices and VMs) is separate and only ~20TB, it has 12x3TB drives, also as two raidz2 vdevs). Both of these servers are 4U and have 24 drive slots.

I have some other servers as well. but they do not have a lot of storage.

1 Like
25

WOW
but that is pretty expensive just for home use … at least this is my opinion. You can never recover the money, but anyway … OK

If anyone does it for the tokens … then you can just invest those thousands of $ for HW to buy the STORJ tokens and you will not waste your time “mining” It seems in your case dealing with HW is a bit of a hobby, so it is good. Do whatever makes you happy.

26

I kin-of did, but that was with both v2 and v3 of Storj. At one point I exchanged a bunch of tokens for quite a few ETH…

I bought the server for v2, but never got to use it. I had a few nodes with single drives for v2 (because starting a new node did not create a loss and, in some cases was actually needed).

Storj v3 requires high reliability, so, instead of using desktops I finished the server and use it with the drives in a zfs pool, because here starting a new node is really bad.

Mining is a bit less risky. Even if Storj project fails and the tokens I still have go to zero, I still have the hardware that I can use for other things. I also need a spare server in case my main one dies down or I need to turn it off for some reason. Mining other coins is different, because bitcoin needs ASICs that are useless without bitcoin and ETH requires video cards and I can only use so many video cards for games. I do mine ETH with some older cards that are still useful for it.

27

That would be me too… I already had pretty much exactly 30TB total before Storj and about 20TB of that was in use. I started with the remaining 10TB and Storj had paid for every HDD upgrade since (3x16TB already), with plenty left to spare. Even though it paid for all those upgrades, they weren’t solely for Storj. My own use has increased as well.

This is clearly not true given what I just said. I never spent a dime on running a Storj node that I didn’t first earn by running a node. It’s definitely profitable and like @Pentium100 said, less risky to only use what you already have or have already earned.

28

I did buy hardware for Storj, though mostly it was for v2. I have a minimum “worth it” level, so, if I think something is going to be profitable, I end up spending money on it to make it “worth it”.

Sometimes I get it back, sometimes not, but if I’m buying hardware and not coins directly, at least I still have the hardware. For example, I bought high endurance SSDs for plotting chia. I still have them and I can use them somewhere else if I need to.

2 Likes
29

Maybe I am not doing thing right. Can you please share your story? How exactly did you earn thousands of $ so you can afford a 30TB server/node? And how much time did it take you to do it?

30

My story:
Create and run some nodes on v2. Earn STORJ tokens.
Create and run a node on v3. Earn STORJ tokens
Wait until 1STORJ price was something like 0.0015ETH.
Sell some thousands of STORJ tokens for a few whole ETH.

In short - run nodes and HODL.

My v3 node gets about $50/month now, but there were various times when Storj either paid multiples of normal earnings (they called it surge pay or something like that) or created a lot of egress for testing, so, at some point I got $100 or more per month in STORJ tokens that were probably worth $0.3 or so.

31

Yes, here are the questions. What did this cost you? And you did not mention how long it took you for everything to payoff. I assume 10+ years? :slight_smile:

32

As mentioned in my previous message, I already owned this. All I bought was 3x16TB to expand.
Which I could easily afford with:
image
I also have 2 smaller nodes which made about $100 each.
I prefer to not even take value increases into account as they aren’t guaranteed, but yeah, it ended up being worth much more than that, since a lot of earnings were made when Storj was close to $0.10.

It also pays being around early, we had a lot of heavy test traffic when there were few nodes on the network, as well as surge payouts up to 500%. Yet even now I make about $60 a month, meaning I can afford a new HDD every few months. Hence why I simply expand when Storj needs more space.

These days it takes longer to get to that point where Storj pays for it’s own space requirements. It really helped that I had 10TB of free space to start out with.
Even today, if you start with 10TB of space, it’ll take a while to fill up, but when it does, you’ve made about $600, which should be plenty to upgrade and have money to spare. The trick is to have a bit space to start out with.

image
image1209×920 114 KB

33

So how many years did it actually take you to pay off just the hdds?

34

Since 2017 for v2 and since 2019 for v3.

My (v3) node dashboard says this:

Total Earned

$2127.60

However, that does not take into account that tokens were cheap in the beginning and got more expensive later. Just by hodling them, I multiplied that amount probably by a factor of 10.

35

That might be my problem, because I think I started in August 2020.
I guess it was really profitable 3 4 years ago. Anyway, for it is still interesting so I will keep going. :slight_smile:

36

I’m not sure the historic info is as interesting as what can be expected currently, but I was able to buy the first HDD after 8 months. But I waited a little longer. Ended up buying the first HDD about a year in, the second 2 months after that. Which was really unfortunate timing as the value of Storj started going up significantly after that. Those HDD’s would now have been worth many thousands of dollars each… oops.

But as mentioned previously, there was heavy test traffic and surge payouts of up to 500% many of those months. That isn’t going to be representative of what you can expect now. The screenshot of the earnings estimator that I posted in my previous message gives a better idea of what you could expect now. If you want to buy a 16TB HDD, you can expect to make enough money for that in about 2 years, if you have at least 8TB available to share now. Of course earnings may still change in the future, but it has gotten a lot more stable since the early days.

I did have v2 nodes before, but I’m not including those earnings as I never made much and I even lost access to one of my old wallets containing some of those tokens. (as a result of a lastpass bug causing loss of data, let that be a warning to not rely on a single place to store your keys. Better yet, don’t rely on third parties at all.)

37

Thank you for the provided information.

38

If you account for storj price going up I made 10k with my nodes this past 3 years. I started with a 3.5TB and a 1TB nodes. So this well over paid all the hardware that I put back into running nodes which was 4 10TB hard drives and 3 rpi4s, and also bought a 3090 with the income I made solely from running storagenodes.

39

If you speculate on the price of StorJ and it’s value going up, then you are always better off purely buying StorJ token directly and you will make more than running a SNO.

40

When I first started running a node I didn’t speculate that the price of storj would skyrocket in the time I was running my nodes. But if I could go back in time I would have baught a whole lot of storj.
I also didnt speculate that the price of eth would be nearly 4000 dollars…
image
This was the price when I first got paid by storj
It was more of a gamble then anything but it paid off and anything I do make from storj now is all profit, If it wasnt for fees it wouldnt be so bad.