I'm creating a kubernetes cluster, and in it I have several services. I know based on https://github.com/GoogleCloudPlatform/kubernetes/blob/master/docs/services.md#discovering-services I have two options.
use the environment variables set by the kubelet.
use skydns
I want to try to use the environment variables first before I go adding another dependency into the mix. However, I'm unsure where the environment variables are for each service. I haven't found them when doing env or sudo env on the kubelet. Are they within a certain container and/or pod? If so do I have to link the other pods to that one to get its environment variables for services?
I have several NodeJS services in containers, so I'm wondering if talking to each service would require this to get the ip:
process.env('SERVICE_X_PUBLIC_IPV4') once I have the environment variable thing sorted out.
Not as important, but related, how does this all work across multiple nodes?
The environment variables for a given service are put in every container that is started after the service was created.
For example, if you create a pod foo and then later a service bar, the pod's containers won't have any environment variables for bar.
If you instead create service bar and then a pod foo, the pod's containers should have environment variables something like:
BAR_PORT=tcp://10.167.240.1:80
BAR_SERVICE_HOST=10.167.240.1
You can test this out by attaching a terminal to one of your containers, as explained here.
Related
I'm using a CI/CD pipeline to deploy a Node.js app on my Kubernetes cluster. Now, we use different sensible environment variables in local, and we would like to deploy them as environment variables within the cluster to be used by the different containers...
Which strategy should I go with?
TIA
There are many tools created in order to let you inject secret into kubernetes safely.
Natively you can use the "Secrets" object: https://kubernetes.io/docs/concepts/configuration/secret/
And mount the secret as env var.
Alternatively, you can use some open source tools that make this process more secured by encrypting the secrets, here are some I recommend:
https://learnk8s.io/kubernetes-secrets-in-git
https://www.vaultproject.io/docs/platform/k8s
Assume that I have a pod active and contains only one active container initially.
This container is a nodejs application in typescript and shows user interface when opened in browser.
Can this container create another container on-demand/dynamically within the SAME POD ?
How can we achieve this? Please advise.
Also, will reusing npm modules like https://www.npmjs.com/package/kubernetes-client help in creating such containers within the same pod?
Can this container create another container on-demand/dynamically within the SAME POD ? How can we achieve this?
No, the containers within a Pod is declared in the PodTemplate that need to be declared upfront before the pod is created. More specific, what use case do you have? What are you trying to achieve?
Also, will reusing npm modules like https://www.npmjs.com/package/kubernetes-client help in creating such containers within the same pod?
A kubernetes client library is useful for interacting with the ApiServer, e.g. for deploying new applications or Pods. But the Kubernetes deployment unit is a Pod - that is the smallest unit you work with. To change a Pod, you create a new one and terminated the previous one.
My app running in a docker container on Azure Webapps for Containers tries to access a connection string through an environment variable. I've added it to the Application Settings in the Azure UI but I can't access it through my code, specifically my ASP.NET Core application is returning null.
I know that the logs won't show it being added as a -e connstring=myconnstring argument in the docker run command, but it should never the less be present in the container.
It turns out, by using the Advanced Tools -> Environment Kudu service in Azure, the connection string environment variable names were being prefixed with SQLAZURECONNSTR_.
I know it is a convention to have these kind of prefixes on environment variables when reading them with the .NET Core environment variable configuration provider as described here, but quite why Azure adds these prefixes automatically, apparently without documenting this behaviour anywhere, is unclear to me.
I have a docker image being deployed from Azure Container Registry. Everything works fine except that I expect to be able to
see environment variables being passed to my running docker image from the Application Settings configured in the portal. It is a Node app accessing environment variables with process.env.VariableName.
As an example its says here
https://blogs.msdn.microsoft.com/waws/2017/09/08/things-you-should-know-web-apps-and-linux/#SetEnvVar …
"App Settings are injected into your app as environment variables at runtime"
I have tried the following.
Setting Applications Settings in the Azure Portal. These are supposed to get passed to the running Docker image as per the documentation but process.env.VariableName in my Node application is not set.
I have tried using a Docker compose file that sets the environment variables but again process.env.VariableName is empty.
I have even updated the VSTS build arguments passing the Variable to my Docker file on build which in turn sets the environment variable. Again no variable passed to the running Docker image.
My conclusion is that custom environment variables are not allowed in Azure Web App for Containers?
Am I doing something wrong?
As per the documentation the app settings are injected into the process as environment variables, however the environment variable name is prefixed with APPSETTING_, so in your Node application you would need to access the app setting with process.env.APPSETTING_VariableName.
I am stuck with terraform. I want to create dynamically ECS services with terraform.
I have a configuration like this :
module/cluster/cluster.tf
module/service/service.tf
What I want to do is inject the service name from jenkins into the terraform configuration, so if the service doesnt exist, it creates it (update it if it exists)
I tried to set up different backend s3 remote state but I don't manage to build the whole infrastructure in one terraform apple.
Is there any way to specify dynamically the service configuration so its create them on demand ?
terraform supports to use variable TF_VAR_<variable> to do on fly change.
From environment variables
Terraform will read environment variables in the form of TF_VAR_name to find the value for a variable. For example, the TF_VAR_access_key variable can be set to set the access_key variable.
Note: Environment variables can only populate string-type variables. List and map type variables must be populated via one of the other mechanisms.
For example,
TF_VAR_environment=development terraform plan
https://www.terraform.io/intro/getting-started/variables.html#from-environment-variables