search

EOA-Blockchain-Labs / ethereumonarm

Tools and scripts to build images that turn ARM devices into ful Ethereum nodes
GNU General Public License v3.0
143 stars 35 forks source link

[Bug] Dashboards not working #58

Closed emlazzarin closed 1 year ago

emlazzarin commented 1 year ago

A fix that also worked for me, from r3dDoX in the Discord:

After checking the prometheus.yml I saw that the configs should all be under: /etc/prometheus/files_sd. This folder is not there on my machine though. So I copied it from the github repo under: fpm-package-builder/utils/ethereumonarm-monitoring-extras/sources/usr/lib/ethereumonarm. And now my Grafana Dashboard seems to slowly get data in 👍

I can look into fixing this at some point, but wanted to put it here in case someone else wants to.

emlazzarin commented 1 year ago

Wondering if the issue is here: https://github.com/diglos/ethereumonarm/blob/b32cc5687137ba0f6851fe2cfdbc266a978cbf0a/fpm-package-builder/utils/ethereumonarm-monitoring-extras/sources/usr/local/sbin/set-ethereunonarm-monitoring-extras#L13

r3dDoX commented 1 year ago

Additionally to the fix you mentioned above, I had two more problems on my dashboards:

  1. On the dashboard 04-EoA Node Status I had a lot of errors saying the duration format was wrong. I saw the use of the variable: $__range_interval in the queries, but I think the interval variable should be: $__interval. So I replaced this variable on the whole dashboard in the JSON model. This seems to work fine now.

  2. Some queries are using the function last_over_time which is not recognized. I researched a bit and saw that this was introduced with prometheus 2.26. But the installed prometheus version seems to be: 2.15.2. I wasn't yet able to fix this one.

r3dDoX commented 1 year ago

@emlazzarin I might have time towards the end of this week. Haven't really read into this repo yet. But I could go from your pointer above and check if I can find out what's happening

washosk commented 1 year ago

Wondering if the issue is here:

https://github.com/diglos/ethereumonarm/blob/b32cc5687137ba0f6851fe2cfdbc266a978cbf0a/fpm-package-builder/utils/ethereumonarm-monitoring-extras/sources/usr/local/sbin/set-ethereunonarm-monitoring-extras#L13

Yes it was the problem, I just upload a new version fixed. Thanks for your help

washosk commented 1 year ago

In the next days we're gonna upgrade Prometheus version to the last stable one, and add new dependencies to dashboards, it should fix everything