Testnet tool

My understanding is that you want to run nodes in a container, with a pre-existing network to connect to.

If you want to do that right now, you’ll have to setup your own network, as Maidsafe don’t have one running publicly. That can be on Digital Ocean, or AWS, or wherever you want to run one. However, for the time being, you need to roll it yourself.

We have a tool that we use for launching testnets on Digital Ocean, and that’s what David is referring to. If you have some technical knowledge, it’s not too hard to use that tool to roll your own network, it’s just that you need to have your own Digital Ocean account. The tool is here.


It’s my assumption that team maidsafe would come with one, but I’m happy to do the “start 11 nodes on one raspberry pi and have other people join me” thing again.

Especially since I still prefer IPv6 over IPv4. I’m not sure if that has changed over the past months/years.

And why would I want to start one on Digital Ocean or Amazon?
I thought the whole idea of maidsafe was not to be reliant on monopolies and olipolies?

The situation just now is very subject to change, so you can’t make a lot of assumptions in the processes you’re going to use.

You need somewhere to host the nodes. Those are 2 possibilities, and we use Digital Ocean for testing. Feel free to host them wherever you like.


I suspect the script they use to create it works with DO, so it was a suggestion to make it easy, as otherwise you’ll have to do more work to set it up.


Yeah, although I think it’s just worth saying that there’s a distinction between being reliant on a service like DO or just using it. The network won’t rely on any given cloud provider for its existence, but I don’t think people would necessarily be opposed to nodes that happened to run on DO, AWS, Azure or whatever. I can only speak for myself, but I wouldn’t be.

We use DO just now because it’s a really useful testing mechanism for the rapid development lifecycle. The same could be said for any cloud provider.


Does DO run the root node?

I think I’ll create two extra project for just being a safe rootnode.

I don’t quite understand your question…do you mean if you’re using the testnet tool?

I’m not sure what you mean when you’re referring to a “project” here either? If this is in relation to your project from the other thread, perhaps we should pick it up in there?

1 Like

The genesis node may well run on a DO-hosted instance to begin with. There are many pragmatic reasons why this is A Good Thing.
The genesis node could run on an AWS node or YOUR hosting, if you could provide the bandwidth and reliability that DO or AWS offer.
We USE Digital Ocean, we are not dependent on them.



Safenet uses a root system for node joining right?
You have a root node, an elder node, an adult node, a child node etc.

Taking a quick look into here
I read that the elder nodes are nodes with decision making powers.

I don’t want DO to have decision making powers.

Github project. Yeah, it’s related to the other thread.

1 Like

DO will NOT have “decision-making” powers. All DO can do is shut down that node if the bills are not paid. In which case another Elder is promoted. No big deal and its my understanding that the use of DO is most likely temporary Also it is likely that the oldest (eldest?) of Elders will at some time fairly soon into the life of the network be hosted on a non-DO instance due to natural churn.


Using the testnet tool, all the nodes run on DO, not just the genesis node.

I would prefer to pick this up in the other thread though, I don’t think it’s appropriate to discuss in here.

1 Like

I am having a poke but I am having trouble connecting to the genesis node. If I figure it out I’ll come back to you.


What genesis node? IP and version in use, please ? :slight_smile:

PS sorry for interrupting if you are having a poke. I hate it when that happens…


Are one of you trying to use the testnet tool? Not exactly sure, but I’m happy to help you get something running if you need any.

There should be a new sn_cli and sn_api version 0.37.0 released tomorrow and the CLI test suite is passing with those using safe_network 0.32.0. Version 0.36.0 of sn_cli/sn_api has a problem with NRS in it, but it’s otherwise usable. Worth just waiting until my PRs are merged tomorrow though to fix NRS.



I have been doing it manually, the last time I tried this worked to connect to the genesis but no luck this time.

RUST_LOG=info $HOME/.safe/node/sn_node --hard-coded-contacts '[""]' --genesis-key b6f6dfb5985145b3f505315106106d8f9b8aae3b20bfb0d2e7xxxxxxxxxxxxxxxxxxxxxxxxx --local-addr --skip-igd

May take you up on helping with the testnet tool :slight_smile:


Sure. The testnet tool will populate a node config for you, which should avoid having to pass that contacts list.


:slight_smile: fantastic news!


I got it working. User error :joy:


@folaht Was this of interest?


This may get long winded and need a thread all unto itself but lets hope not :slight_smile:

My first roadblock is

Initializing the backend...

Successfully configured the backend "s3"! Terraform will automatically
use this backend unless the backend configuration changes.
Error refreshing state: AccessDenied: Access Denied
	status code: 403, request id: WCEKQVPSNBMAKE75, host id: MQPPMTwgpbKa5pl