I think it would be nice to have a thread where we can share how many nodes are successfully running on specific hardware and internet connectivity. Something like that:
OS
Number of nodes
Internet connection speed
Processor (CPU architecture and the core count)
Ram
Hard drive
optional - estimate of power consumption and local electricity costs
Let the discussions be elsewhere and here is just a list for easy browsing
Internet connection speed = 80Mb/s Down, 20Mb/s Up
Processor = ARM (Raspberry Pi4)
CPU x Core count = 1 x 4 = 4
RAM = 8GB
Hard drive = 256GB M2 SSD dedicated to safe dirs. OS on the SD card.
OS = Ubuntu 22.04 LTS
safe-site_2:-
Number of nodes = 2
Internet connection speed = 80Mb/s Down, 20Mb/s Up
Processor = ARM (Raspberry Pi4)
CPU x Core count = 1 x 4 = 4
RAM = 8GB
Hard drive = 128GB SD Card the OS is on. Sceptical if this is a good idea but it’s a allegedly high endurance one. We’ll see how it works out.
OS = Ubuntu 22.04 LTS
Notes - The crappy BT router in this location starts to grind to a halt when there are more than a few hundred connections open so 4 nodes is the absolute limit and I don’t want to be called up all the time so 2 is the sensible limit.
Edit
Added OS.
Edit 2
Added Core count
Edit 3
Oops! It’s not an NVMe drive - Just a M2 SSD.
Seems a good approach although I personally won’t touch Google Docs unless my life depends upon it. I remember suggesting we do something like this many years ago and it feels great to see us reach the point where it is needed for real.
Does that mean it can also be wiped out and intentionally deleted? or others overwriting other’s rows?
Hmmm. Not that folks have bad intentions here but yeah, we need an easier way to account for this stuff if folks really want to keep a friendly record off potential hardware specs and # of nodes etc.
I think the beta participants early on already answered a bunch of these questions, but the summary and analytics off that data have yet to be presented or summarized by Maidsafe (if even planned).
On another note, I can see individual posts on this thread also spiraling out of control.
Easier to have a website to submit from drop down multiple options per set of node configurations and aggregrate to a central DB, yet at same time have it summarize and searchable.
It even be more interesting if a parameter in safenode passed on telemetry data (optional) to endpoints that Maidsafe hosts to collect all this data during beta phase and beyond. It be more systematic, accurate, and organized in the backend, prior to any frontend dashboard created for the generic audience.
After a laptop failure stopped me for almost two weeks I got back to working on my demo just today, and while it isn’t Google Docs, it will be interesting to see if people begin to use it for things like this and - if things carry over to the final network - begin creating the first things on Autonomi that will be there forever.
That’s an even more exciting prospect to me than being among the first wave of beta, though I’ll be doing both.
Sure, I can connect my power consumption metter (from eth classic mine days) to the socket, but my i7 will surely consume much more power than r pi for instance.
And how i can calculate power consumption for vps?
Let say that I will have to upgrade vps surely with storage and let’s take that in to a count and say that vps is costing 25 eur/month that would be 0,31 eur / node per month or 3,72 eur / node / year and you have nothing in house .
That’s the point. If you are leaving your laptop on to run nodes 24x7 that period of extra use is worth knowing. If you only run them while doing other work as normal, it will be marginal and can be ignored.
No need, because the cost of this is factored into their charges.
It’s just to help give a more complete comparison between approaches.