Grafana "Gitlab Overview" board doesn't work

Observability
1

Hi,

Our Gitlab instance is installed on a dedicated, in-house server, and Gitlab, which is currently the only application on this server, is installed using the Docker omnibus image.

We’ve just upgraded from the 11.10 version to the latest 12.1.1, and I enabled the Grafana boardmore out of curiosity than need. The various builtin dashboards were available, and all seemed to work except for the Overview one, see attached screenshot.

image
image1916×770 60.2 KB

I find it a bit ironic that the uptime would work in the specific service boards but not the main one :wink:

Our instance is behind a reverse-proxy which handles the https unpacking.

Has anyone ever seen something like that, or has any idea where I could get more info ? I can’t find proper logs for Grafana, i’m afraid!

2

Maybe too late, I saw the same screen recently.
No data in overview, chart for redis, nginx and so on shows metrics data.

In my case, the reason was that node-exporter is disabled.
(node-exporter is disabled by default for docker, by the commit below.)

Enabling node-exporter with GITLAB_OMNIBUS_CONFIG env var solved the issue.

P.S. The author says The node_exporter requires access to the host, so it doesn't make sense to run it in the application image.,
but I think some of the metrics is still available and useful.

3

Maybe too late, but you can try this below:

add these config to gitlab.rb:

## Enable node_exporter
node_exporter['enable'] = true
node_exporter['flags'] = {
  'collector.textfile.directory' => "/var/opt/gitlab/node-exporter/textfile_collector",
  'path.procfs' => "/host/proc",
  'path.sysfs' => "/host/sys",
  'path.rootfs' => "/rootfs",
  'collector.filesystem.ignored-mount-points' => "'^/(sys|proc|dev|host|etc|rootfs/var/lib/docker/containers|rootfs/var/lib/docker/overlay2|rootfs/run/docker/netns|rootfs/var/lib/docker/aufs)($$|/)'"
}

add these volume mappings to docker-compose.yml:

services:
  gitlab:
    volumes:
      - '/proc:/host/proc:ro'
      - '/sys:/host/sys:ro'
      - '/:/rootfs:ro'

The solution is by enable the node_exporter and also mount different directories from host, while using startup flags to specify right dirs for node_exporter to run properly.
I’m facing the same issue and finally I’ve figured it out. Hope this can help more!