Testnet tool

Yeah so that’s why you’re not getting any logs, obviously because the node isn’t running. I don’t know at what point the testnet tool starts up the node, but I assume it’s during the provisioning stage. The output from Terraform may indicate why it’s not starting up.

1 Like

I think it’s been mentioned before, but we really do need an easy way of telling whether our nodes are running or have crashed.

Do you mean nodes you’ve launched on your own machine?

Yes. I mean the health of a node on my own machine. Stuff like ps aux | grep safe is not enough.

get-info returns
line 21: rg: command not found
status also tell me that get-info: line 21: rg: command not found

get-info line 21:
VILLAGE=$(rg “Becoming Adult|Becoming Elder” logs/* -c | wc -l)

That may be true, but we’re not going to get something to address that for a while. I would doubt before a launch candidate.

You would need to write a separate piece of software to sit in a user’s tray or something like that (and you’re also dealing with cross platform complications here) to monitor the service, and that’s not a completely trivial thing.

Once we have a working network we could look at bells and whistles like that, but for now it would be a low priority in my opinion.

1 Like

It looks like you need an installation of ripgrep for using this script.

1 Like

Ok thanks.
I could pull all the logs without it complaining about aws. :partying_face:

1 Like

Cool. ELK would make all this pain go away :).


I think if @happybeing or someone else can resurrect vdash then this itch can get scratched. That tool shows a lot of potential if it can be made current and extended. Its way out of my Rust skills to help though…

I would contend that working on vdash would be a BGF-compatible project and there could be some reward for those with the skills to help.

1 Like

ELK is causing me pain right now with logstash consuming ~200% of CPU on the VM constantly right now. Seems I am not alone in seeing this though.

If I am understanding things correctly in order to use a specific sn_node version I need to either have a local copy of it or as per line:57 up.sh a copy on aws or elsewhere.
Is it possible to install a specific version directly from github using the tool?

Also having difficulty with netdata, are there any specific changes I need to make to get it working.

At the moment, any version that’s there’s a Github release for, should also be on S3. Part of the release process is to upload to S3. That’s a public bucket, so there should be no problem with pulling from there.

Sorry what is ‘netdata’?

Hey, how are you getting on with the ELK setup?

If you’re having problems with local resources we can get some VMs spun up on the Maidsafe DO account for you to play around with and make some progress with it.

On a project I was working on last year, we actually ran ELK just fine on Kubernetes. DO have hosted Kubernetes instances, so we can use that. If you want to just get your head around ELK first though, we can get a few VMs to work with.


When I try to use put-random-data I get a netdata error. I am guessing it has to do with this but see no reference to it anywhere but the error.

Not working for me, will delve deeper.

1 Like

I think you may just be getting a confusing error message back from the CLI here.

What exactly does it say?

What specific version are you trying to use?

Will let you know when I am back at pc, will be a while though.

0.32.0 not a big deal, using a local copy works fine.

1 Like

That version is definitely in the S3 bucket, I just downloaded it. Let me know what the specfic output you’re seeing is.

1 Like

Off the top of my head “aws not found” but will confirm.