Community Test 13 November - offline

I have a pi node myself that I could maybe persuade a pal to stick on his network. If that works I would have three nodes, house, AWS and a pi round at Dougies…
I also have theoretically some spare sites with SiteGround but so far I have only used that for WP hosting - let me look at that…


I’ll do what I can but I might (will) need a little hand holding.

If you set this up for me so it’s plug and play I’ll happily buy the fish supper and pay postage.

1 Like

I’ll set up a pi4 and post it down to you - in theory should just need powered on and your home wifi should assign it a local address. Might need some port forwarding footering on your router. What model do you have?

1 Like

Standard talk talk router :woozy_face:

Send a picture of the back of it with the model number.

1 Like

Device ver: FAST 5364-3. T8

err it seems the pi4 is spoken for already …

but I also have a couple of old banana pis that worked OK a couple of years back. I will drag them out and see if I have not repurposed the SD card with the image. Advantage of the banana pi is that it has a SATA slot so you can plug in a std HDD - no faffing with USB attached storage. Disadvantage is its not the fastest but we are looking for proof-of-concept here.


I do have a laptop but it’s windows 10 and I’m not sure if it’s up to it?

Suck it and see?

There are windows binaries…

Anyway you could boot it from an Ubuntu USB stick/SD card. Which is easier to post than a Pi + PSU

If your offering and happy to do the hand holding I’m in :+1:t2:


I’ll need to find a boot image for the banana Pi. As I suspected the SD card got used for something else Or I send you an Ubuntu live disk and a script to set it up.

1 Like

Whatever is easiest for you (and me) to get us going. If it’s easier just let me know what I need to buy so it will work better than windows.

1 Like

Buy a Pi4 and load Ubuntu on it

Maybe we should design a plug and play node and make the image publicly available?

This isn’t the cheepest by a long shot but it’s what I bought for my dedicated node.

Seagate BarraCuda Q5 2TB PCIe NVME M.2 SSD £183.93 Ebuyer
Raspberry Pi Compute Module 4 (CM4) 8GB Lite £61.99 RRS
Tofu Switch Blade Enclosure £27.34 Oratek
Tofu Carrier board for Raspberry cm4 £77.33 Oratek
Tofu M.2 Mkey adapter £14.84 Oratek
Total £365.43


UFW uncomplicated firewall or the GUI version GUFW.

Still not sure if any firewall is needed if its a pi or a spare old machine with no other use exposed to the wild and wonderful world.

Are you sure that we cannot run multiple nodes on separate devices from the same IP?
I have run 2 pi and a pc from home simultaneously before.

I was just coming to post that link.
If we run say 15 instances at $48 per month with the aim of trying to keep something up almost permanently that is $720 per month or $23 per day if we can achieve something being up pretty much 100% of the time.
Is it worth aiming for 100% up time, probably not yet.

Is it more reasonable to have 1 test per week which is $23 for each day that it runs and gives development time between tests, perhaps instead of weekly a new test with each new release?

This is why I went with the instance one size down, at half the cost it is only $11.50 per day (for 15 instances) which is no train wreck if it is only a few days here and there!
But even that is $360 per month if we push for constantly having something up.


AIUI we need ONE big droplet that can run multiple nodes from and then all us punters connect from home. We dont need a 4vcpu8gb each.

Were these connected to a test network or just a network on your LAN? If it was an actual Maidsafe test network then I’m wrong and this becomes a LOT simpler.

1 Like

One droplet running multiple nodes? not sure that I follow.

yes to a test network.
I smell good reason for a new test!


I thought you had multiple nodes running from the big droplet attached to multiple (virtual) network interfaces?
Or am I getting this completely wrong with the notion that each node needs its own IP address?

More and more Im thinking I got this wrong and it becomes much simpler and cheaper.
Did you kick off 17 or so nodes from one IP address on a big DO droplet? How come there was 17 seperate IPs in the logs? I know some other folk connected but the vast majority of the nodes particpating were DO instances. Or at least thats what I thought.

1 Like

No each node has its own droplet, hence my reluctance to use the 4vcp-8gb droplets without good reason it adds up rather quick. So far the 2vcp-4gb droplets appear to work ok.


If we can do it cheap - all from home or ONE $48/mo DO droplet then we should aim for 100% uptime. And when we get enough participation we will have 100% uptime even without the big DO instance(s).

EDIT: The cunning plan to use SiteGround failed - it was a long shot anyway - they are set up purely for hosting, its not like an AWS instance…

baseos |$ safe node install
   0: Error fetching list of releases for maidsafe/safe_network repository
   1: IoError: no /proc/self/exe available. Is /proc mounted?
   2: no /proc/self/exe available. Is /proc mounted?


Backtrace omitted.
Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
1 Like