Canonical Observability Stack Lite

Contents:

• Introduction
• Configure MicroK8s
• Deploy the COS Lite bundle
• Deploy the COS Lite bundle with overlays
• Browse dashboards
• Next steps
• Further reading

Introduction

The COS Lite bundle is a Juju-based observability stack, running on Kubernetes. The bundle consists of Prometheus, Loki, Alertmanager and Grafana.

Let’s go and deploy that bundle!

Configure MicroK8s

For the COS Lite bundle deployment to go smoothly, make sure the following MicroK8s addons are enabled: dns, hostpath-storage and metallb.

You can check this with microk8s status, and if any are missing, enable them with

microk8s enable dns 
microk8s enable hostpath-storage

The bundle comes with Traefik to provide ingress, for which the metallb addon should be enabled:

IPADDR=$(ip -4 -j route get 2.2.2.2 | jq -r '.[] | .prefsrc')
microk8s enable metallb:$IPADDR-$IPADDR

To wait for all the addons to be rolled out,

microk8s kubectl rollout status deployments/hostpath-provisioner -n kube-system -w
microk8s kubectl rollout status deployments/coredns -n kube-system -w
microk8s kubectl rollout status daemonset.apps/speaker -n metallb-system -w

Deploy the COS Lite bundle

It is usually a good idea to create a dedicated model for the COS Lite bundle. So let’s do just that and call the new model cos:

juju add-model cos
juju switch cos

Next, deploy the bundle with:

juju deploy cos-lite --trust

Now you can sit back and watch the deployment take place:

watch --color juju status --color --relations

The status of your deployment should eventually be very similar to the following:

> juju status --relations
Model  Controller  Cloud/Region        Version  SLA          Timestamp
cos    charm-dev   microk8s/localhost  2.9.42   unsupported  15:43:36-04:00

App           Version  Status  Scale  Charm             Channel  Rev  Address         Exposed  Message
alertmanager  0.25.0   active      1  alertmanager-k8s  edge      67  10.152.183.93   no       
catalogue              active      1  catalogue-k8s     edge      15  10.152.183.193  no       
grafana       9.2.1    active      1  grafana-k8s       edge      77  10.152.183.137  no       
loki          2.7.4    active      1  loki-k8s          edge      82  10.152.183.119  no       
prometheus    2.42.0   active      1  prometheus-k8s    edge     122  10.152.183.51   no       
traefik       2.9.6    active      1  traefik-k8s       edge     125  10.43.8.34      no       

Unit             Workload  Agent  Address     Ports  Message
alertmanager/0*  active    idle   10.1.55.34         
catalogue/0*     active    idle   10.1.55.38         
grafana/0*       active    idle   10.1.55.32         
loki/0*          active    idle   10.1.55.14         
prometheus/0*    active    idle   10.1.55.40         
traefik/0*       active    idle   10.1.55.53         

Relation provider                   Requirer                     Interface              Type     Message
alertmanager:alerting               loki:alertmanager            alertmanager_dispatch  regular  
alertmanager:alerting               prometheus:alertmanager      alertmanager_dispatch  regular  
alertmanager:grafana-dashboard      grafana:grafana-dashboard    grafana_dashboard      regular  
alertmanager:grafana-source         grafana:grafana-source       grafana_datasource     regular  
alertmanager:replicas               alertmanager:replicas        alertmanager_replica   peer     
alertmanager:self-metrics-endpoint  prometheus:metrics-endpoint  prometheus_scrape      regular  
catalogue:catalogue                 alertmanager:catalogue       catalogue              regular  
catalogue:catalogue                 grafana:catalogue            catalogue              regular  
catalogue:catalogue                 prometheus:catalogue         catalogue              regular  
grafana:grafana                     grafana:grafana              grafana_peers          peer     
grafana:metrics-endpoint            prometheus:metrics-endpoint  prometheus_scrape      regular  
loki:grafana-dashboard              grafana:grafana-dashboard    grafana_dashboard      regular  
loki:grafana-source                 grafana:grafana-source       grafana_datasource     regular  
loki:metrics-endpoint               prometheus:metrics-endpoint  prometheus_scrape      regular  
prometheus:grafana-dashboard        grafana:grafana-dashboard    grafana_dashboard      regular  
prometheus:grafana-source           grafana:grafana-source       grafana_datasource     regular  
prometheus:prometheus-peers         prometheus:prometheus-peers  prometheus_peers       peer     
traefik:ingress                     alertmanager:ingress         ingress                regular  
traefik:ingress                     catalogue:ingress            ingress                regular  
traefik:ingress-per-unit            loki:ingress                 ingress_per_unit       regular  
traefik:ingress-per-unit            prometheus:ingress           ingress_per_unit       regular  
traefik:metrics-endpoint            prometheus:metrics-endpoint  prometheus_scrape      regular  
traefik:traefik-route               grafana:ingress              traefik_route          regular  

Now COS Lite is good to go: you can relate software with it to begin the monitoring!

Alternatively, you may want to deploy the bundle with one or more of our readily available overlays, which is what we’ll cover next.

Deploy the COS Lite bundle with overlays

An overlay is a set of model-specific modifications that avoid repetitive overhead in setting up bundles like COS Lite.

Specifically, we offer the following overlays:

• the offers overlay makes your COS model ready for cross-model relations

• the storage-small overlay applies some defaults for the various storages used by the COS Lite components.

To use any of the overlays above, you need to include an --overlay argument per overlay (applied in order):

curl -L https://raw.githubusercontent.com/canonical/cos-lite-bundle/main/overlays/offers-overlay.yaml -O
curl -L https://raw.githubusercontent.com/canonical/cos-lite-bundle/main/overlays/storage-small-overlay.yaml -O

juju deploy cos-lite \
  --trust \
  --overlay ./offers-overlay.yaml \
  --overlay ./storage-small-overlay.yaml

Browse dashboards

When all the charms are deployed you can head over to browse their built-in web-UIs. You can find out their addresses from the show-proxied-endpoints traefik action. For example:

juju run traefik/0 show-proxied-endpoints --format=yaml \
  | yq '."traefik/0".results."proxied-endpoints"' \
  | jq

…should return output similar to:

{
  "prometheus/0": {
    "url": "http://10.43.8.34:80/cos-prometheus-0"
  },
  "loki/0": {
    "url": "http://10.43.8.34:80/cos-loki-0"
  },
  "catalogue": {
    "url": "http://10.43.8.34:80/cos-catalogue"
  },
  "alertmanager": {
    "url": "http://10.43.8.34:80/cos-alertmanager"
  }
}

In the output above,

• 10.43.8.34 is traefik’s IP address.
• Applications that are ingresses “per app”, such as alertmanager, are accessible via the model-app path (i.e. http://10.43.8.34:80/cos-alertmanager).
• Applications that are ingresses “per unit”, such as loki, are accessible via the model-app-unit path (i.e. http://10.43.8.34:80/cos-loki-0).

Note that Grafana does not appear in the list. Currently, to obtain Grafana’s proxied endpoint you would need to look at catalogue’s relation data directly - try running:

juju show-unit catalogue/0 | grep url

…which should return a list of the endpoints like this:

      url: http://10.43.8.34:80/cos-catalogue
      url: http://10.43.8.34/cos-grafana
      url: http://10.43.8.34:80/cos-prometheus-0
      url: http://10.43.8.34:80/cos-alertmanager

With ingress in place, you can still access the workloads via pod IPs, but you will need to include the original port, as well as the ingress path. For example:

curl 10.1.55.34:9093/cos-alertmanager/-/ready

juju run grafana/leader get-admin-password --model cos

Enjoy!

Next steps

• Use the scrape target charm to have the COS stack scrape any Prometheus compatible metrics target.
• Relate your own charm to the COS stack with relation interfaces such as prometheus_scrape.
• Configure alertmanager to have alerts routed to your receivers.
• Use the cos-proxy machine charm to observe LMA-enabled machine charms.
• Use the grafana-agent machine charm to observe charms on other substrates than Kubernetes.

If you need support, the charmhub community is the best place to get all your questions answered and get in touch with the community.

Further reading

• Model-driven observability: modern monitoring with Juju