Hi,
I would like to share my dashboard too. Monitoring is realy a cool part of a stake pool project.
http://dashboard.qcada.com:3000
Grafana wiht Prometheus is realy powerfull, i will certainly give a try with GitLab, Apache or VMware!
I don’t have metrics about Blocks because QcADA is a baby pool for now
Here are variables for two nodes (block_producer and relay1) in prometheus.yml
static_configs:
- targets: ['block_producer_IP:9100']
labels:
alias: 'core'
type: 'core'
- targets: ['relay1_IP:9100']
labels:
alias: 'relay1'
type: 'cardano-node'
- targets: ['block_producer_IP:12798']
labels:
alias: 'block-producer-node'
type: 'core'
- targets: ['relay1_IP:12798']
labels:
alias: 'relay-node1'
type: 'cardano-node'
To download : https://www.qcada.com/share/grafana-dashboard-QcADA.json
Comments or suggestions are welcome
Luc [QcADA]
2 Likes
Hi,
With the new update 1.25.1, metrics name changed and Prometheus can’t find the data.
Any idea ?
Yes, update ur panels in grafana with cardano_node_metrics
For example
cardano_node_metrics_remainingKESPeriods_int
Or
cardano_node_metrics_txsProcessedNum_int
Yeah, i tried but still no data.
Prometheus show the old nomenclature.
Adrem
25
Scroll further down and you’ll see the ones with metrics as a prefix.
Still no data.
I opened port 9090 for testing.
http://3.97.117.186:9090
Not that way… Choose a pannel click on edit and send me the screnshot i will tell u what to modify
Ok, delete chainDB.
From now on should be just cardano_node_metrics_xxxx
For epoch will be cardano_node_metrics_epoch_int{…
Still no data, i gona recheck configs.
Thanks for help
Refresh!
I configured : cardano_node_metrics_epoch_int and working
Adrem
32
+1 to what @Alexd1985 mentioned
Check your queries for correct syntax and compare to 1.25.1 release notes on github
delete {…} and add it to legend tab
or try cardano_node_metric_ChainDB_epoch_int{…
Adrem
36
This may be a silly question, but have you checked server-side that your nodes are running happily after the update?
Adrem
37
And yes, the syntax you posted in the pic is correct.
PS do you get data for the other metrics or are they all giving you grief?
PPS are you also using node_exporter for machine metrics? Are those live?
Then… u have another issue, try another pannel …like tx or peers
and shouldn’t be prometheus with small p as source?
Nodes are ok, gLiveView give right data (TX, peers, KES).
All services are Running. prometheus-node-exporter on BP, prometheus.service, grafana-server and prometheus-node-exporter on Relay.
The hardware metrics works fine, like node_load1{alias=“core”}
Show the pannel for load1
type curl localhost:12798/metrics | grep epoch