Hello all,
>From the docs (https://prometheus.io/docs/guides/dockerswarm/) you can
easily use the Docker UNIX Socket like so:
- job_name: 'docker'
dockerswarm_sd_configs:
- host: unix:///var/run/docker.sock
However I'm struggling to work with a Docker API that is exposed over HTTP
with a custom port and path. For example:
- job_name: 'docker'
dockerswarm_sd_configs:
- host: http://docker_proxy:8000/proxy
This doesn't work and results in a dns resolution failure.
I've also tried relabel_configs like:
- job_name: 'docker'
dockerswarm_sd_configs:
- host: http://docker_proxy
role: nodes
relabel_configs:
- source_labels: [__address__]
regex: ".*"
target_label: __address__
replacement: $1:8000/proxy
But this doesn't work either, the error suggests that the port isn'g being
used to connect to the Docker API at all, nor the path.
I must be doing something wrong, or using a custom port + path isn't
supproted?
The reason btw I'm doing this is because the upstream prom/prometheus image
(*rightfully so*) now runs as the "nobody" user making it quite difficult
to bind-mount the Docker UNIX socket in a consistent way. So running a
proxy that re-exposes it over HTTP and drops privileges seems to be the
best approach here.
Kind regards
James
--
You received this message because you are subscribed to the Google Groups
"Prometheus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/prometheus-users/82c16918-b027-4e75-b2c3-910fe7b4037bn%40googlegroups.com.
