Excessive use of hdd on my node

Start my node on May 15, 2020 with a capacity of 3Tb today, June 10, 2020, the node has consumed 1Tb of Hdd with hardly generating any benefit.
Is this normal? Which may be due ? is there a problem with my node?

It takes months to generate some profit based on the time it takes to ‘vet’ your node, then the held payments. You can look at one of the online calculators for a general idea of what to expect.

1 Like

Actually, getting 1TB of data in your first month is very good. Usually, the first month is only vetting period, which doesn’t get you lots of data.

Just keep going, payouts will ramp up after some months.

Increasing storage every 15 days for a simple promise to collect “something” after a few months … is a very bad business…

It’s not just a promise and you can calculate what you receive (of course this might still fluctuate to user demand because it is not mining!).
But you definitely receive at least 1.5$/TB/month. However, for the first month you have a held back amount of 75% so you would only receive 0.375$/TB/month (assuming that your node has 1TB for the whole month, which is of course not the case in the first month of operating a node).
It will get a lot better in consecutive month because the held amount decreases and your stored data increases.

I didn’t even calculate the egress payout in, which could be around 10% of what you have stored so e.g. 100GB which would be 2$

What do you mean by “Increasing storage”? Does someone force you to add more HDDs every 2nd week?

1 Like

To me, it’s a very good business. If you don want the cash and if you don’t have a bit of fun setting up this whole thing, then it might be bad business for you, though.

You must be of letters. Mathematics is not your forte.

Ok. Would you still please elaborate what you mean by “increasing storage every 15 days”?

Please, if you have a point or a retort then go for it.
Posting just to criticise someone’s skills is pointless and does not forward the discussion.


It’s a fair business. And you shouldn’t want anything else. Fair compensation leads to a sustainable business. If you would get paid more than it’s worth you can never count on it sticking around. That said, yesterday I purchased my second 16TB HDD just from my STORJ earnings. I actually exchanged the tokens for BTC and paid for that. The nice bump in STORJ value put me right over the edge. I started out with sharing 10TB of spare storage, now (due to loss for redundancy and personal use) I will be able to share a total of 28TB after that HDD is added to the array. Storj is more than paying for itself at this point.

And you seem to be the opposite, nobody understands what you’re talking about.

It might be useful to expand on this. It also helps to either quote the post you are referring to or to hit the reply button on that post.

There have been a couple of crazy months, like January where I earned $50 in one month from a 500GB hard drive, but even with the current rate (which I would consider normal for the network), an 8TB hard drive would pay for itself in a year. Assuming a disk will live for 2-6 years, that’s good profit.

The network is very “balanced”. You can’t get an insane amount of money by adding a large amount of storage, it has to fill up slowly first. The holding back money system also ensures people are in it for the long haul. It may not seem worth it at first, but if you look back after a year or two you will realize it’s actually surprisingly profitable.


lol well i kinda see what @TheMacR is saying… i’m 4 months in now and previously had 3 months on a v2 node… which i stopped because i needed space and wanted to come back for v3, kinda misunderstood that v2 didn’t get shutdown…

so i’m like 7 months in now… with my present system i got like 48tb worth of drives used mainly for the storagenode and redundancy, the node has always gotten whatever space it could take…

totally thus far i’ve been paid out 20-25$
so yeah… i know it’s mainly my own fault for misunderstanding that v2 didn’t get shutdown.
but thus far v3 has been only expenses, i still cannot cover the base electricity costs.

but i’m happy to hear that should change xD tho not sure i should expect the same … i mean 50$ pr 500gb that would put my node at 240$ a month, so 2880$ in a year.
that would most certainly put my ROI inside of two years, even tho it’s mostly old junk i’m working with… ofc if i was to add my own hourly wages then… well not sure why i’m doing this lol

If you need $100/TiB/month for a ROI in 2 years you’re definitely using too expensive hardware.

From my experience I usually get $4/TiB/month. For an 8TB drive (~7TiB) that’s ~$30/month. An 8TB drive costs 160-170 euros here (a WD My Book / Elements enclosure), ~$190. Ignoring held back amount, that’s ROI within 6 months!

Even if you needed to buy server hardware (~$50 for raspberry pi 4 with power supply) and include electricity costs $.20/kwh → $20-40/year.

One hard drive and one raspberry pi, one time investment of around $250 and ~$30/year running costs. Payout around $300 for a year. I’d say that’s pretty profitable: profit over 3 years would probably be $500-$700. Depending on your job you might not think this is worth it, but I think it sure is!

ofc if i was to add my own hourly wages then… well not sure why i’m doing this lol

Assuming you are a basic sysadmin (i.e. you know what SSH is, how to set up a webserver or samba share) it should take you 2-3 hours max. In normal conditions, the storage node needs no maintenance, other than a docker(-compose) pull like your other software.

1 Like

i decided to go the route where it’s more economical as i scale… also here the electricity prices are more like 0.30$
and i kinda wanted a server anyways… kinda want to build a small data center, and only 4 months in… and system is only at 10tb… i should be able to keep in profit at 1.5$ per tb when i pass the 100-150tb range.

so yeah i would have been better off with a rpi, but thats like a toy with no real HA features, going with a server should give me a lot more stability… even if i current is still in the red.

and setting up the system… well i decided to make the switch from windows to linux using zfs … so lots of things to learn and mistake to make and suffer for… but long term i’m sure it will be a great benefit… one i’m paying for in the beginning.

I have a relatively large server, Ryzen 7 1700, 32GB RAM, many hard drives, I use it for a LOT of stuff. Still, one 8TB storage node manages to cover electricity costs for the entire machine. I think that’s pretty impressive, by adding one hard drive I have removed yearly running costs of my server!

It uses 70-100W depending on load, even if your electricity costs were as high as 35 cents it’s $200 in electricity compared to $300 payout from the storagenode

i’m on an old dual cpu x5630 (16threads) rack server, with 48gb ram… can’t really get it below 220watts continual draw… doesn’t seem to affect it if its running full tilt or 1% cpu utilization…

i do really enjoy having my own server tho, also set aside a 10x10meter room to make a proper server room. xD

ill be happy if it just covers the electricity, but ofc after 4 months i’m still getting like 30% payout… getting close to earning what i kinda suspect my electricity costs are… sadly i will need a DAS next time i expand my pool… or replacing existing smaller drives with some 16tb or so… but that gives me the trouble of getting rid of the hdd’s again… but we will see what happens … for now the system isn’t going to change to much before the node is getting close to full…

zfs also really loves to have many drives… so many drives lol

how much wattage do your ryzen draw?

I don’t know the CPU specifically, but the entire system with 3 SSDs, 7 HDDs, 2x16GB ECC memory and a ryzen 7 1700 draws 70W-110W depending on load.

I used to have a server with an x5650 and moved to Ryzen. Ryzen is insanely cheap, very performant (way better than the x5650), power efficient, and has ECC support if you choose the right motherboard.

ZFS is great but I wouldn’t use it for storagenodes. You lose a lot of capacity to parity and metadata. Running one storagenode per disk seems most profitable and is the recommended setup

yeah, but i think i paid 5$ pr cpu… so difficult to beat that lol 10$ for 16 threads…
but yeah you really save on the power there…

i kinda like running it on zfs… tho after having moved a 6tb and now 9tb node a couple of times between pools… well then i really would have liked to do it correctly from the get go… really the whole plan with having a big pool is that i can share the iops between many tasks, and get the advantages of arc, l2arc and slog on everything.
and then there is the redundancy thing… which already saved my pool a couple of times because the server has been mistreated and thus has some issues with bad drives… maybe they didn’t like being exposed to minus 20 celcius last winter xD

i kinda like my old junk server… allows me to be a little more fast and loose with my choices on how i maintain the system… recently learned that controlling humidity is pretty important… lol literally everything is corroded inside it… at one point the fan power cables started breaking by themselves because they had corroded through lol

but yeah had i taken into account the power costs long term… then i might not have made the choices that i did… on the other hand then i might have had a almost brand new totally corroded server by now … i suppose there are always advantages and disadvantages to everything… :smiley: