Oh got’cha!
so that’s how You see it?
You think its free?
ITS NOT FREE!
Equipment under the strain of STORJ, is wearing out.
Pay for maintaining nodes fairly.
Got a whooping 1 dollar(!) for entire month of 07.2023 from traffic, of a full 7TB node.
Right now, one may think,
that it would be better, to just leave that additional space unused.
3 Likes
Something is wrong with your numbers. One of my nodes has 1TB filled and 20GB egress and I’ll also get 1$.
You get 1.5 USD per TB for storage alone. If your node is full, you’ll get 10$ already. Was it not running the whole month?
Okay, yes, my node is full, BUT what with all those who just got e.g. only 1TB out of 7TB filled?
and need to wait 2-3 years to fill that HDD to the full?
What if the HDD dies before that.
What if it doesn’t? A HDD can die anytime, with or without storj, or it can run for 10 years without problems. That is called risk assessment. Leave your space unused if the risk is too high for you.
Storj doesn’t make any promises how long it takes to fill a node. That is just the false expectation of newcomers. 2-3 years is probably to low, but that’s how it is.
Make advertising for storj such that they get more customers and you get more data.
I don’t get why people complain so much. Storj is basically passive income. Not much, but you don’t have to do much either. After setup it’s basically fire and forget, just an occasionally linux maintenance.
There is just one rule: Use the hardware you already have. Want to do an investment and buy new stuff? That’s on you, don’t complain if it doesn’t work out!
I feel this is going off topic.
9 Likes
You kidding? yea sure, HDD can die anytime, but when it dies because of nodes intensive usage, that’s no problem for STORJ, because it’s free resource like Alexey stated, and storj don’t care, i guess we’re just suckers, letting ourselves be taken advantage of to a group of people, who don’t care about original ideas, just pretend… and they doing it badly. I will do whatever i want, dont worry about me, im telling in general, what STORJ is doing is very bad for SNOs and the future of public network, because they don’t want to do the reforms. it has no future for SNOs longterm as it currently looks. I wrote how it can be done better for the benefit of all, but all I hear in response from the STORj admins are excuses that this can’t be done, or that that can’t be done, or that the transfers people have are too small, in fact, which I refute once pore as much, but it doesn’t work on them. There is simply no will to do anything, for SNOs, because it is not their pain, that SNOs equipment bleeds for them for semi free, after pay rate change, the traffic did not rise. But they are fine, plenty of tokens for theirs salary’s. Now you can see why, they don’t want to make efforts, when they can’t get good customers without certificates anyway, hence all these excuses, better to take care of where the money is, sure I understand that, just don’t say you can’t do anything, just tell the truth, that you don’t want to, and that’s it. And the fate of SNO is worse than indifferent to you, in case of anything, you have a lot of people willing to hop in our place from centralized datacenters. Its fine. Everything is fine.
BOY, “You don’t have to do much” ? on what world You are living BOY?
just read the forum, constant problem solving, im not telling its bad or wrong, but certainly not “don’t have to do much”, and "passive income, yea sure, pasive, until its NOT, when things get broken You have to fix it asap or You risking getting DQ or suspended.
Also who am i to make ads for STORj? well i did fomr writings few times, letters to some company’s. You? just curious.
Its going oftopic, because you directed it to such tracks, talking nonsense: No one said investing. Stop making things out.
I said DESTROYING currently owned equipment, for not enough, to even replace it, when it will break, under the load of STORJ usage. I dont know what HDDs You have, by mine dont break out of being IDLE! So one could be better of not joing as a SNO at all. Im fine with that risk, BUT i don’t think that’s what STORJ should offer to SNOs, STORJ can do much better terms for everyone, customers, SNOs, and even make bigger earnings for them selfs! I showed how!
1 Like
Isn’t that what you signed up for? Of course HDDs break faster when in use compared to being idle. What do you expect?
Yes sometimes there are problems, but that is not constant. 99% of the time everything runs fine.
My nodes were offline for a week when I didn’t have monitoring and I wasn’t getting DQ or suspended. Just last week I had to fsck one HDD which took 4 days. The node went from 94% to 84% Online score. I think you are over exaggerating.
No. I use it personally and for my customers as well. Nothing big though.
It was more a general comment. Sometimes one has to speak his mind. Like you did.
It’s always good to put links to what you are talking about.
1 Like
If the drive dies due to mechanical reasons it’s because its main bearing wears out. Not due to “excessive” motion of an arm with a head assembly. So once your drive is spinning it’s slowly moving to the grave. Whether you have it read or write during this time is more or less irrelevant.
Are you saying the measly IOPs that storj inflicts on your array is “destroying” your hardware and performance? You need to configure your array better then. Personally I don’t notice the node running at all, on any of my appliances.
And then, if additional IOPs load mattered — did not you buy drives to use them? Do you want to throw them away in pristine condition? If you don’t use them — why do you have them?
What are you taking about? Storj is paying me real money for nothing. They are literary charity from my perspective. In what world would it be appropriate to complain about the amount of free money gifted?
LOL. First of all, there is a strong selection bias. I’d say bias is too soft a word. It’s selection exclusion. Nobody posts daily “today nothing happened everything works as expected” topics for you to draw conclusions. And 100% of problems I read about on this forum are self-inflicted by SNOs who are unfamiliar with what’s involved with running services 24/7 and the storagenode is their first service they trying to ramp up on. Of course they make rookie mistakes, can’t configure router, surprised by the UPS and filesystem quirks.
So, to counterweight the noise, here is me claiming that it took me few hours to setup the node (because I wanted to use FreeBSD and storj does not have a service definitions, so I wrote a script to set it up and keep updated. I could use docker and avoid all that initial time investment altogether) and since I’ve started it — I have not touched it. I don’t monitor it either, besides uptime robot. It’s been a year. Maybe more. During this time I started two more nodes with the same script, and they also demanded zero maintenance. Storj team does all the work and pushes updates. I just pocket 100% free money every month for doing nothing. Running storj does not cost me anything because I can read and follow advice and only use underutilized resources. I already had maintenance free 24/7 running home servers. Adding storj did not change maintenance requirements.
I honestly spent much more time figuring out how to sell tokens than keeping node running.
So, piggibacking, on what planet are you living that your nodes require constant babysitting and that you complain about effortless passive income that it is not large enough?
Very bizarre comment.
5 Likes
Ok, you have two choices: do not share excess storage and bandwidth and get nothing in return, still paying your bills, or you may share it and get a discount to your bills. Always your choice!
We do not suggest to invest in a first place and always told to beginners to use what they have now and what’s already online with Storj or without, if the Operator decided to made investments - it’s their choice and their risk.
1 Like
Depends on what state you are referring by IDLE? Slowdown? Spinddown? Headpark? What drive model?
There are thousands of possibilities. I can even think of one where STORJ could help prolong your HDDs life. Let’s assume you have a WD Green and you self host Nextcloud on it. Now you read and write small Word files from it. The WD Green pretty aggressively parks the heads in between your usage. That will wear down the HDD quick and killed a lot of WD Greens. Now thanks to STORJ, your drive is always busy, will never head park. You just got a longer lifetime thanks to STORJ 
I know this is not a very common scenario, but HDDs mostly wear down because of power on time. Power on time does not change because of STORJ.
1 Like
I got a disk thats dying… think its my 8th since i started Storj 5+ years ago…
have sort of stopped keeping track… but maybe i should start keeping a proper log on that.
the disks i stop using for storj can usually work for less high workload tasks for years afterwards.
but they will start throwing errors when doing 24/7 workloads.
duno why exactly… maybe a heat thing…
is rather interesting… because they run fine in other zfs pools for years afterwards.
my media server is made from 3 x 6TB disks that came out of my Storj setup.
anyways…
to calculate your ROI just assumed 1.6$ pr TBm thats about what you will get these days.
and yes you will need a lot of data… but hey look at the bright side… the less full the disk is, the less active it will also be… so less wear…
also enterprise disks come with a 5 year warranty… and with good reason…
i’ve had one of my 18TB HDD’s be nearly DOA.
i will agree with @Ruskiem, it is very hard for me to see how free hardware resources is suppose to be used, when the Storj workload is so high…
i’ve tried using my Storj hardware for other things… and it haven’t gone very well, because of the often semi high latency on the Storj hardware.
so i just stopped… but yeah it could use spare disks… that would be sitting around anyways…
but then they could just be cold spares… with storj they earn a bit… and it does cover the cost of the hardware… atleast when they are filling up nicely…
but that can take a long time.
Could not agree more. HDD are dead after 5 years. I write of all tech to 0$ after the warranty runs out. Everything on top is extra. A 1000$ laptop with 2y warranty is more expensive than a 2000$ laptop with a 5y warranty.
That is also why you should not care, you use your unused disk space within your warranty period. Everything else is extra.
2 Likes
Even without any investment, using only what you have… still running Storj nodes cost you money, because the drives work 24/7, and consumes electricity and produce heat; the switches and routers consumes electricity for each I/O from Storj. If you don’t run nodes, the drives goes to sleep or hybernation, routers and switches are calmer etc. Even the fans speed lower, because no more heat from drives… So running Storj nodes can’t be free in any scenario. This leting aside the harware dieing sooner and so on.
2 Likes
Can you quantify what the difference is in terms of cost such that we can establish once and for all if it’s basically free or not?
They do anyway. Maybe the use 0.01 W more because of the head movement.
the switches and routers consumes electricity
Again, they run anyway. I don’t shut down my router during the night, nor my Wifi.
What do you think is the power consumption difference between an idling switch and a switch transporting 50mbit/s?
the drives goes to sleep or hybernation
Nope. Two reasons. First, wear out is higher, second ZFS defaults.
routers and switches are calmer
My SFP+ switch and my opnsense 10Gbit Router both run passive. If I can run 10Gbit passive, a normal 1Gbit most definitely is also passive.
So running Storj nodes can’t be free in any scenario
Ok, free is maybe a little bit over the top. Let’s be more realistic. Let’s say 2W extra? 7$ a year if you pay 40cents per kWh.
harware dieing sooner
Again, not true necessarily.
2 Likes
Sure.
It has a fan, but that fan only gets active if you put lot of power hungry SFP+ into it. Something stupid like 10GBASE adapter. I have 3 10Gbit DAC and one 1Gbit RJ45, fan only runs during rare reboots.
PC is just an old Lenovo SFF, some old i5 8th gen Intel CPU, unplugged the CPU fan and the PSU fan. CPU runs at around 70°, sometimes 75° under load. Power mode was changed from Adaptive to Hiadaptive, otherwise single connections speeds were capped at 6Gbit.
For the network card, I use a used Chelsio T520 with an attached Noctua 40mm fan (not audible) because I was to cheap to buy something passive like Intel or Mellanox. That fan also cools the the PSU by pulling air right next to it.
1 Like
I use RB4011 at home, it role all my network. when i tested storj client all my 500 mbit trafic made only something like 10% of CPU.
I hate those drives so much. When I get those drives in used systems normally I just toss them in the trash.
4 Likes
I would be very cautious using Mikrotik for switching.
Had a few of them only only had problems, had random storms in the network, interfaces stopping passing data, interfaces passing only some data etc.
Now running Cisco for switching (both Catalysts and Nexuses) and have zero issues.
For switching, honestly even TP-Link is better than Mikrotik.
For routing, if you are willing to pay to be a beta tester with the new v7, go ahead, but ideally buy x86 license and build your own hardware, so you will have an option to switch to something else, as their routers, even the very expensive ones, are usually hit and miss if you are doing something serious with them - as for example the constant Kernel panics on CCR1072, phantom broadcast storms on P2P links on CCR2004 and CCR2116, prematurely dying PSUs and leaking caps on previous CCR generation and so on.
And every software update is a drama and doing that remotely without OOB access is a suicide, as in every software release something gets broken, even a simple stuff as SFP interfaces.
I manage a bunch of CRS switches on an ISP network and they seem to work great. The only downside is spotty compatibility with 100M SFP modules, but otherwise they are great.
Both the last v6 and v7 work great for switching.
I know of a CCR where the power supply caps failed. I do not know how old it is, but it’s old. Better yet, I could just buy a new power supply and replace it (it was faster than replacing caps on the old one).
As for Cisco - I really, really hate artificial limits. If the manufacturer sells me a device and then expects me to pay more to use it (port licenses etc) I will do my best to avoid buying such device.
Mikrotik for routing and switching, Ubiquiti for WiFi APs is a great combination in my experience.
The only time I had problems updating Mikrotik firmware was when I tried to update from old v6.x to v7. Updating to more recent v6.x and then to latest v7 works OK.
Maybe I am not using them “seriously” enough though. most of the time it’s just a few vlans and a couple of ACLs, so far I did not try to have 400 vlans on it, maybe that makes it crash.
–
Speaking about Storj profitability, to me, there has to be enough income to make it work spending time running the node. Yes, in theory, if I could use unused resources and get $10/year it would still be profit, but adding the time needed and the transfer/exchnange fees and the $10 disappears.
4 Likes
If all SNOs whould used only the leftover space, retired surveillance HDDs, SSDs and 2.5" HDDs in their laptops, etc. the network whould never hit the proportions that has today. The big majority of PCs and laptops today have SSDs (all flavors), that are not suited for Storj. Laptop drives die verry quiqly, so again, they are a no no. You are left with NAS and servers, and grandpa PCs. How many of them are hobbists, not interested in money? How many of them loose interest or need that space after a while, producing repairs and increase the costs for Storj? Even Storj team, using those big payouts, incentivised the investitions in new hardware, even though they keep that lame excuse “we just tell you to use what you have… bla bla”.
If you show me money, I’ll do my best to take them. If you don’t want complaints, don’t show me the money!
You want a reliable network, enough space for small and big clients, than aproach the problem taking into consideration that this network need good drives, needs investments from SNOs and needs payouts accordingly.
From every angle you aproach the “use what you have” mantra, you realise that’s just stupid and a lame excuse for Storj to not be accounted responsible.
3 Likes