Hi, a stupid question.
I have 2 nodes om the same submet and there is a third node running on the same subnet.
I have the possibility to change the subnet for both of my nodes and make sure that I get one with no other node, but does it make sense?
Is it better to have 2/3rds of traffic of one subnet or get 100% of traffic of one subnet, knowing that next to me there is another subnet, so effectively receiving 50% of the two subnets?
I hope the question makes sense.
Thanks
It depends, why are you here? to make money or just for fun.
You will fill Your disk 2 times slower if sharing ip with some other node, so even if You for experimenting and fun stuff, its always better to have no neighbors imo.
I suddenly got a neighbor once and noticed my ingress dropped immediately. I rebooted my router to get a new IP, and the problem was solved.
Now I run the following Bash script daily to check for neighbors:
my_ip=$(curl -s https://ipinfo.io/ip)
neighbors=$(curl -s storjnet.info/api/neighbors/${my_ip} | jq .result.count)
if [ ${neighbors} -gt 1 ]
then
echo Storj: ${neighbors} neighbors found!
exit 1
else
echo I am alone
fi
My NAS will send me the result if unsuccesfull.
Brgds
Chasing neighbors is pointless.
Storj is a cost-recovery mechanism for idle, already-online capacity. Keywords being “Cost recovery” and “Already online”. If your setup economics depend on squeezing subnet placement, you’ve already violated the premise.
It’s important to understand that the urge to optimize is designed into the system, and is a gamification artifact: A metric exists, so people try to maximize it, and when the number moves on a graph they get a tiny dopamine hit.
Ingress charts, vetting, neighbors tracking turned a background utility into a Skinner box. So now grown adults reboot routers to roll a new IP, pretending this matters let alone worth their time. This is quite ridicuolous, let aline for $10/month.
Metric chasing is not rational. It’s a waste of time. There is no meaningful objective – why are you doing it?
If you need to micromanage neighbors, you’re not sharing “idle capacity.” You’re running a tiny, unprofitable business badly, all while pretending it’s a hobby.
Before you start scripting router reboots, some retrospection is warranted here:
Running storagenode is not a competitive sport. Turn off the graph. Don’t look at the dashboard. Watched kettle never boils. Let the disk fill when it fills. You get some electricity costs offset. Not everything needs to be, let alone is worth being, optimized.
If you change subnets, aka instead of aaa.bbb.ccc.ddd, you move to aaa.bbb.xxx.ddd, you will have full traffic splitting between your 2 nodes.
Now you just get 2/3 of the uploads from clients.
I don’t understand were’s the confusion?
The geo location matters, not the consecutiveness of the IP addresses, if you were reffering to that, for winning races, but you don’t know were the other guy is, even if it’s on your subnet now.
Who knows how the ISP allocated the IPs. But the sattelite dosen’t realy care. It dosen’t send traffic split between nodes gattered in one geolocation if they are on different subnets. It dosen’t do geolocation for this purpose.
But I thing you missunderstood what is a /24 subnet. To be on a different subnet means to have one or more of the 3 first groups different, aka aaa or bbb or ccc.
If you and the other guy have the first 3 groups identical, you are on the same subnet with him. Changing the forth group of numbers dosen’t help. It’s the same /24 subnet.
That part was correct, then - water. If you had stuck to your thoughts yourself, you wouldn’t even have written a line of text here.
If you can change something for the better (more profitable in this case) and it doesn’t cost you anything, why shouldn’t you do it?
Because it does not cost anything if your time is free. Compulsively checking dashboard, writing, debugging, and maintaining scripts to reset DHCP in this case, it’s all not free. It takes resources and attention. Every time you think of storj you don’t think about something more important.
It’s not free. It’s very expensive. Now you can read my comment above having this perspective in mind.
Look at this this way - if that was so important – there would not be neighbors. the other guy would move out from that subnet. But they don’t. Why? Because they understand it’s irrelevant. Storj still pays regardless of whether there are 3 nodes on the subnet or 30. It’s still free money.
Ok, I get it. I looked at your profile and I dread to think how much Storj pays you for being on this forum 
hehe 
I just like ranting on forums
Indeed, let me add a bit more to that.
I have an 11 TB node running on a single dynamic IP. This is just a hobby for me, and I treat my node like a plant—taking care of it and watching it grow slowly. Halving my ingress would ruin my goal of reaching 12 TB.
Writing the script took me 10 minutes, and resetting my router took just one second. I’ve only had to do this once in the 3+ years my node has been running.
Cheers!
This purist approach woudl be admirable if true, but according to the rest of your message this is absolutely not what you are dong.
Instead you could do any number of the following:
If being alone on the subnet woudl be important – the other node woudl leave. You don’t have to do anything.
This woudl be quiet brillanace. Instead, you do hands-on payout optimization, narrated as gentle cultivation.
Hi, speaking for myself: I treat this as a fun activity where I can tinker a bit. Given that the rewards are not virtual points of some kind but actual money (more or less) gives me more motivation.
Optimizing neighbors is part of the tinkering I like to do, certainly not in order to justify running storj as an investment!
I woupd be interested in your points about minimizing time to forst byte and reducing HDD power consumption, do you have good resources that explain those points?