To start the containers, use the below command
docker-compose up -d
To bring down the containers, use the command
docker-compose down
mkdir /tmp/k3dvol/postgresdata
mkdir /tmp/k3dvol/dashboards
cp monitoring/grafana/provisioning/dashboard/*.json /tmp/k3dvol/dashboards
k3d cluster create todo-cluster -p 4200:80@loadbalancer --api-port 6550 --volume /tmp/k3dvol:/tmp/k3dvol --servers 1 --agents 1
kubectl config use-context k3d-todo-cluster
k3d cluster delete todo-cluster
Folling values need to be set before installing the helm charts
slackWebhook
invalues.yaml
client-id
andclient-secret
in0021-todo-secret.yaml
client-id
in0033-todo-ui-secret.yaml
helm install todo ./todo-chart --namespace default --wait --timeout 300s
helm uninstall todo
Now that our pods are running, we have the option to use the Prometheus dashboard right from our local machine. This is done by using the following command:
kubectl port-forward -n monitoring prometheus 9090
Now visit http://127.0.0.1:9090 to access the Prometheus dashboard. Otherwise, you can access the dashboard via http://localhost:4200/prometheus
You can visit http://localhost:4200/alerts to access Alert Manager dashboard.
Visit http://127.0.0.1:4200/grafana to access Grafana dashboards. Use the admin
/password
combination to login.
In current context, when you access your Angular app from browser, you're using the Angular app in host machine, in host network, not inside container anymore, your Angular app is exposed by mapping port from container to host machine. Therefore when you make an API request, it'll coming from host machine -> docker container, at that time, host machine will try to resolve service name in current context and of course from host machine there's no service name
Therefore, for all browser apps, even you deploy it using docker container, but when you call API, mostly you'll have to use localhost:<mapped_port> or something that resolve to host machine not inside docker container (Eg: domain name).