can't work with solo address
Closed this issue ยท 4 comments
Hello,
Thanks for the report! Are you running it locally or under Kubernetes?
Issue is likely in the way your Prometheus source is configured, the model is:
- the watcher exposes its Prometheus metrics link on port 8000
- Prometheus polls data from the watcher periodically, it needs to be configured to do so, if you are running under Kubernetes with the provided chart it should be automatic (the chart exposes the metrics port as 8000)
- is your Prometheus source properly setup on Grafana?
I'm replicating a local setup to improve local setups documentation.
thanks for your feedback,yes,i run it locally and i do miss the configuration for Prometheus.I add it now and some of the pannel show the data now.Appreciate for that.
Also,if it possible to provide more detail about loki information? don't know how to configure it but Grafana require for the data source from loki
What I did to have it work locally:
- install promtail/loki
- configure it to watch a folder where log files are stored, I've added the following extra-entry in
/etc/loki/promtail.yaml
underscrape_configs
:
scrape_configs:
- job_name: system
static_configs:
- targets:
- localhost
labels:
job: varlogs
__path__: /var/log/*log
- targets:
- localhost
labels:
app: ethereum-validator-watcher
__path__: /var/log/ethereum-watcher/*log
Then I make sure my user can write to this log directory:
sudo mkdir -p /var/log/ethereum-watcher
sudo chown :mxs /var/log/ethereum-watcher
sudo chmod 775 /var/log/ethereum-watcher
Then when I run the watcher, I redirect logs to the directory as follows:
eth-validator-watcher --beacon-url $RPC_CL 2>&1 | tee -a /var/log/ethereum-watcher/watcher.log
It's clearly a bit tricky to setup, I'll probably find a way to have this under docker with everything ready directly maybe to simplify local setups.
make it works now,thanks a lot