It makes me question the legitimacy of the network size calculation.
2 Likes
It makes me question the last 12 months if I’m honest.
5 Likes
It makes me question how quickly I can make RPi4 Node boxes and push them into friend’s and family’s houses!
I had thought the best thing to do was to give people 4GB RPi4s in a nice heatsink case with a 500GB SSD with the OS on a 32GB MicroSSD. That can run 10 fully provisioned nodes with ease. The cost is about £150.
But in one house I have supplied just a RPi4 in a basic case with just a 128GB MicroSSD for the OS and Node storage. And it’s worked without fault through thick and thin for this network through testing, abuse and production usage. I didn’t expect the MicroSSD to survive long even though it is allegedly ‘High Endurance’. But it has. The total cost is less than £100.
I’m beginning to think that is the way forward. I realise the MicroSSD will fail at some point but then so can any of the components of the expensive option.
If network bandwidth is the current biggest concern in the endless battle between that and network connections, compute, memory, storage and storage performance then cheap hardware in general and a lot of cheap nodes in a lot of locations seems to be the flavour of the month.
I can put one of those a week into someone’s house. The only issue is I’m running out of friends I can risk annoying by their home network being disrupted and my time to manage the boxes. But if I can guarantee there will be no impact by the network being predictable and keeping the number of nodes low I could deploy more.
Given recent events I’m starting to think a lot of fully provisioned cheap RPi4s running about 5 fully provisioned Nodes is better than a smaller number of ‘better’ boxes running 10 fully provisioned nodes.
9 Likes
Hey @Mightyfool
The spike from yesterday has definitely settled.
It might be interesting if you started uploading again and we can see what happens.
I would ask people here, if possible, don’t decrease your fleet size during these uploads. The spike was correlated with a reduction in network size of about 500 nodes, so it would be good to see what happens without a size reduction, if that would be possible, or maybe how long we could sustain uploads without a size reduction.
Mighty, if you do start, let us know what time.
5 Likes
For anyone following along here. I just discussed with Chriso and I’m slowly ramping up my upgrades, starting now.
3 Likes
For anyone appreciating these efforts and they want to help cover some of my cost. I’ve opened a donation address:
Removed
I am happy to contribute to the project wherever I can and will continue to as long as I can, but I do need to keep an eye out for my own financial position. So anything that could help cover some of the cost is greatly appreciated.
There is also a donation address for the Seed Phrase Hunt I initiated earlier last month. Given the difficulty of the planned hunt, it would be amazing to see that get some support as well. 🔐 Seed Phrase Hunt 2.0 – The Long Game
2 Likes
Just for some context from 1 person who needs to keep reducing his fleet.
My issue is I have lost ssh access to some of my machines so they are powered off until I can sit and connect a monitor to each individually which is a major pita that I can’t currently find the time for.
But also as others reduce numbers my available disc space is taking a hit so I need to reduce too.
Vicious circle, bigger ssd’s and restoring access is on the roadmap..
65 bigger ssd’s is a major flaw in my setup though 
, somebody pump that ant please 
3 Likes
No worries Josh.
It was just if it was possible, but I understand and appreciate the resource you’re providing.
5 Likes
I am cautiously increasing in my house and those of friends. Fibre upgrade on Tuesday. If all goes well I’ll be able to boost my own contribution to the network. And another friend should be coming online next week as well.
@Mightyfool I’ll bung you some ANT for uploads and the prize fund. I have a feeling this will be like the early days of Bitcoin and people were throwing around values for fun that were worth pennies then but would buy a house a few years later. But this is all ‘ill gotten gains’ and the values are entirely notional until the network flies.
5 Likes
Ubuntu running 10 nodes (no upnp) , network traffic continues to steeply climb (~10Gb per node used). At some point my 80/40 link will get saturated.
1 Like
I’ll provide 47GB of uploads from us (or as far as the gas I currently have in the dev wallet will take me), starting now.
6 Likes
I just won the lottery, found the old keys on a backup 
7 Likes
And my 10 node machine, now 15 nodes is back up to 3Mbytes/s to 5Mbytes/s
Its at the tipping point of needing to reduce nodes again. (Just saw a quick peak of 11Mbytes/s which is allowed by the router on this 9-10MBytes/s uplink using some of its GB of buffer)
2 Likes
The first 0.001 ANT payments went out.
On average nodes are now at 12 GB storage used, putting us on the purple dot in the whitepaper graph:
Small disclaimer, according to the whitepaper, price is determined by amount of chunks stored / total chunks storage available so there could be a small discrepancy. I’m not sure if anyone here knows what the total amount of chunks is a node can hold?
4 Likes
Nodes can store 16K records max
The price will be based on records stored and not amount of bytes.
In theory a node storing only 4MB records will fill up to 64GB
32GB was chosen as an average mix of record sizes and average number of records a node will be storing during normal network operations in a global environment*
* meaning global adoption
ALSO we were clearly told the curve is basically unit-less and as a indication of the pricing. Yes made plainly clear when I tried to get more details on it, like the algorithm and units
6 Likes
Thanks for the feedback.
On our 200 nodes, I’m seeing some increase in CPU usage, but there haven’t been significant effects yet on bandwidth.
2 Likes
Joined to 3dp club today too! 
0.00162873108471456
6 Likes
How many nodes?
1 Like
1 Like
Hey @Mightyfool
I’ve stopped our uploads. Got through about half of them.
From our metrics I could see we were under quite a bit of strain, though it wasn’t as pronounced as the spike from a few days ago. I was just a little bit concerned about continuing and then losing another batch of hundreds of nodes.
I should have gathered some useful data I can use. Thanks for your help.
I would probably advise you to take it easy. I know we would all be keen to see the ANT price going up, but it may be better if we can get things in better shape first.
9 Likes