Storj node easy mass deployment solution

Full compose multinode deployment for storj-node with exporters and grafana, ideal for mass node deployment.

This is a compilation for multiple solutions, put together in one simple to use package.
For detail on all solutions and credits, check out github. It would not be possible without them.

storj-deployment on GitHub

Node features:

  • storj-node with log to file
  • logrotate for storj-node
  • preconfigured zksync
  • storj exporter
  • storj log exporter
  • storj updater
  • running on shared dedicated internal network

Host and UI:

  • docker host OS monitoring
  • prometheus + grafana configuration
  • grafana storj and host os monitoring graphs
  • simple website for host displaying all available web interfaces
  • storj multinode optional deployment

Tutorial video
Tutorial covers everything from docker instalation to deployment and configuration.
Included full closed captions in case you have problem understanding my accent and transcript of all used commands in comments.

UI Preview
Grafana storj node statistics (storj-exporter and storj-log-exporter)

Grafana docker statistics (host-exporter)

List of web interfaces

Deployment preview

storj-node stack

storj-ui stack


I would also like to add monitoring and alert system. Something that would at least send an email to warn about issues.
I was thinking of sensu, since I have experience in large scale with it, but that may be an overkill.
Does anybody have any experience/suggestions with such monitoring for storj/docker?

You can actually set it within grafana.
I have it alert me in discord whenever a node has a problem.

that would be great. could you please point me in the right direction? I’m new to grafana.

thank you. I will look into it and add it to deployment.

1 Like

I like messaging on telegram with uptime-kuma.

yea, I plan to look into multiple solutions to have more options
not sure what would be the best yet
mail notifications are basic and I certainly want those. then different type of instant notification for serious issues, like node crashing due to disk not being available and such.

grafana alerting seems to be moved from graphs to dedicated section which does not have option to import/export.
so this will be a no-go, at least for now

I believe you can add this type to the existing Prometheus too.

there may be a way to do this using grafana provisioning.
this is new to me, so may take some researching and lots of testing.