heliumdiy/helium_ansible

ansible playbooks for managing helium hotspots

helium_ansible's Issues

Rename latest-arm64 in gateway-latest

https://quay.io/repository/team-helium/miner?tab=tags&tag=latest

rename

helium_ansible/group_vars/all.yaml

target_miner_tag: "latest-arm64"

in

target_miner_tag: "gateway-latest"

to install the current miner firmware - is it sufficient to enter the change in all.yaml?

Disk cleanup

• Setup docker log rotation parameters
• Cleanup blockchain db when disk % reaches a certain threshold.

Allow multiple vpn providers

Enable the usage of tailscale and openvpn for example.

Ansible tailscale

Ansible wireguard

Metrics Server (Dashboard) #TODO

• create Ansible role for a metrics server (prometheus, loki, prompushgateway, grafana)
• create shell script with functions to query api values and submit them to the Prometheus Push Gateway
• Create ansible tasks for the shell script with variables for hotspot addresses or account addresses to extract data for

Remove useless conditionals from the tasks

With #11 we no longer need to have some conditionals in the tasks. So they can be removed.

Write documentation / howto guides

Ansible pushgateway

Include in instructions the need to set "enable_packet_forwarder" to True in all.yaml settings.

After following the instructions to setup on a fresh Rak V2, it took me several hours of debugging to realise I needed to edit the all.yaml file to enable packet forwarding before pushing to the device.

I think it's not an unusual usecase to have the device do the packet forwarding, so maybe add a step in the instructions to say to do so if you aren't using an external packet forwarder.

New gateway_v2 hung? If so this is a fix..

I have been running the ansible for a while the new latest 1/29/2022 update has a loading issue or gets “hung “ if that is the case for anyone running this chmod 755 is as per the tasks… I ran a chmod 777 on miner_data and it started actually putting files into the folder. 🤷‍♀️ Just a hint if yours gets stuck. Or “hung” but either way be patient with it , it takes a loooooong time to load all the stuff I’m over 30 mins with it.

Replace Promtail with Grafana Agent

Replace promtail with grafana agent with the exact same logs configuration to allow a lightweight metrics shipping to a prometheus remote-write endpoint.

Readme for this repo

Provide some examples on the readme on how this can be executed against one, or an entire fleet of miners.

Provide example on Group Vars and Host Vars

People should be able to clearly see what this inventory does, and how does the structure of the repo align in respect of the structure of the groups.

• Setup groups for each region
• Fix inventory example
• Decide if we need to create groups for hotspot vendors or if this should remain in host_vars.