Oh I see, thanks for the tip! I’ll check it out
from experience i can tell you it is indeed not worth it to run a node on it.
For me it was just to see if it was possible, would work and to kill some old harddisks which would otherwise just be trashed anyways.
At least they are now used until death, and were useful for a bit.
Hey everyone! Just wanted to let you know i have been working on a similear project to show people the economics of storj nodes. I’ve got a dashboard that offers public access to real-time stats, including data usage, earnings, power costs, and more. It’s been great monitoring everything in one place, and if you want to take a look at how it’s performing live, feel free to check it out: http://www.hgsoftware.dk
Please go to:
pistorjnode
for the stats.
i also recently released a new video explaining the dashboard.
Hope its fine to share here.
Peace out, HG
12 Likes
I completely forgot my promise to dodge a laser, and that really, really bothers me now.
That being said, I made a new video where I turned my ridiculous project into a project that’s just slightly ridiculous. Cheers to that! 
8 Likes
5 posts were split to a new topic: Improvements on video about the nodes setup
Guys, i did it today,!
I saved a nice server case with X10SLL-F and Lsi9211-4i from recycling (for free!)
With 8 bays and no disks, sadly also with only 4GB ram.
I beginn to gather more specs, and maybe in a couple of months it wil join the network.
Any suggestions/tips ? installing an OS would be the next step.
1 Like
4 GB is fine, I run 3x14 TiB disks without an issue with 6 GB ram
free -h
total used free shared buff/cache available
Mem: 5.8Gi 718Mi 111Mi 0.0Ki 5.0Gi 4.8Gi
Swap: 4.0Gi 854Mi 3.2Gi
1 Like
Add more ram (ram is cheap on ebay), install TrueNAS Core, enjoy über reliable NAS, use my scripts to run nodes in jails.
im already on ebay searching ram 
1 Like
three separate nodes or one big one?
If writes start getting backed up (test onslaught combined with slower storage) then sometimes the storj node software can start gobbling up ram and cause an out of memory situation. I’ve had a single (very slow performing) node use 4GB RAM. But under normal ‘light’ conditions it’s vastly less.
Each HDD is a separate node.
MOtherboard: Supermicro X8SAX - The Retro Web
Using a 16 bay SuperMicro case CSE-836
Model name: Intel(R) Core(TM) i7 CPU 960 @ 3.20GHz
Stepping: 5
CPU MHz: 3200.023
2 Likes
19 posts were split to a new topic: Power consumption of the old hardware, prices across the globe
How are the four regular cases mounted in the rack?
@Vadim , sorry for pinging you, but I know you use Toshiba MG10 drives, and I’m planing to build some rigs for Storj, without UPS. How are they in production? Did you have any problems with power outages? Are they too slow for winning races or the performance is good enough? Any advises?
For anyone reading: Toshiba MG10 drive has a cash for safe keeping data unwritten to disk in case of power outages, but only in 512e format.
Thanks!
I like them, and use them everywhere. I do not have problems with a at all. From all of my disks only 1 hdd had problems after 3 years, and was changed by warranty, as they have 5y.
I do not have other disks with that size, so cant tell are they faster or slower than others.
But when you have so many small files all disks are very slow.
Also they cheaper that other disks. for me 20tb is around 320EUR
1 Like
I use them too and I love them. They are fast enough to win races (I have 99% win). So my biggest impact is more the connection speed than hard drive.
And I like that they are cheap and have good quality
I have 70+ of those Toshibas (16…20TB). 5 have failed so far, all after approximately 3 years. BTW: This is not in line with the bathtub theory… 
How loud are those Toshibas? I do not mean the dB level from the spec (32 dB). More if the sound from them are annoying.
Not much difference to Seagate or WD enterprise HDDs. Those parts are not made for a living room.
I believe if you would have 70 Exos, the same or more would fail in the same time.
From what big storage companies are publishing it’s also hard to make a good fail rate prediction.
I have 15 Exos 16TB+ and, since 2021, none have failed. I also have 2 IronWolf 8TB, they are the oldest in my farm. They are still kicking in the lowest RAM setup possible, so every walker is crushing them with no mercy, and they keep going hard, with the same earnings like their big brothers.
Fail rate is verry low this days, so I won’t take that into account.
The main concern now is to make setups without UPS, because UPS is too costly, takes to much space and can fail without notice. It’s better not to trust them.
I don’t know if WD has the same persistent cache as Toshiba, but they are too expensive anyways. So if Toshiba is performant enough for nodes, and it’s the cheaper option, I will go with those from now on.