I was planning to use Jenkins(CI) ----> Spinnaker(CD) integration for AWS EKS.
Does Spinnaker support multi-cluster deployments?
For example:
I will have 4 clusters in different accounts
and I want to have 1 Spinnaker deployed to one of the clusters and manage other 3 as well.
is it possible to do?
Yes is possible. The suggested way is to have Spinnaker Running in an AWS Account called Spinnaker or CD and in a specific namespace called spin.
A great guide to follow for Spinnaker in EKS is Continuous Delivery using Spinnaker on Amazon EKS
Related
EMR uses Yarn but there is an option to run on AKS with k8s as resource manager. What about serverless EMR, it seems that the application looks like a multimaster k8s deployment so I would assume that k8s is the resource manager, any pointers?
thank,
Mike
architectural question
How to create an IAM role to run workloads on Amazon EMR on EKS?
The official documentation remains very vague on this point in particular
This is multistep process. We have EKS cluster on which we run EMR jobs. So we need to setup IAM + EKS role mapping configurations. AWS Documentation is detailed which creates confusion sometimes.
Following are high level steps which might help you to refer AWS documentation again and understand.
Note: These steps assumes you already have created EKS.
Create Namespace to run your spark jobs.
Create RBAC Role and Role binding for your cluster configuration. (This is EKS level role based access mechanism.)
Edit aws-auth to update roleARN for AWSServiceRoleForAmazonEMRContainers. (EKS Authorisation)
Create EMR virtual cluster and assigned it to EKS namespace created above.
Create trust policy to access EKS cluster for EMR containers.
Create Job execution role and associate above policy to it.
Submit your emr on eks job.
I was planning on installing Istio on my new AKS cluster. However, in the prerequisites for Istio, it is mentioned that K8S cluster should have RBAC enabled. However, I read that AKS ( preview ) doesn't have it enabled. Is this true? Is there an option for me to try Istio on AKS.
AKS is GA and looks like RBAC is available now,
https://azure.microsoft.com/en-us/blog/azure-kubernetes-service-aks-ga-new-regions-new-features-new-productivity/
In fact RBAC is not available in Azure AKS currently. According to this GitHub issue it is on the roadmap for Q1 2018.
In Azure you can use ACS which is an older version of AKS, but with control over Kubernetes master or acs-engine where you have full control over Kubernetes cluster.
AKS now enables RBAC by default.
There are also docs on how to install Istio:
https://learn.microsoft.com/en-us/azure/aks/istio-install
We are doing POC for service mesh on our AKS cluster using istio. I have found a very good guide to install istio with all its components on AKS cluster and it does not require any RBAC on AKS. Infact this guide is cloud agnostic. I am not sure if it production graded ready but working like a charm till now. Just apply first 3 files and 4th one optional. The name might be little confusing for you. But its working on AKS very well. Hope that worsk for you.
Istio Installation Files
kubectl apply -f 1-istio-init.yaml 2-istio-minikube.yaml 3-kiali-secret.yaml
I was trying to deploy AWS services using cloudFormation. I was successful with deploy for particular region. Now i wanted to deploy some of AWS Services in different region for example i have EC2, Lambda and S3 for deployment and i have to deploy EC2 and lambda on us-west region and S3 on EU-East and US-WEST region.
Can this possible with one template.
I went thought AWS Stack Set but i think this will deploy to all AWS Service to all mention region. I wanted to have some AWS Services to some region and some with only one specific region.
Assuming you're using the CLI, your best option is to have multiple profiles configured and then perform two deployments with different profiles for each deployment. Secondarily, you can use parameters as input to your template and use a conditional statement to deploy different resources based on the region you're targeting. Relevant links -
http://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/parameters-section-structure.html
https://forums.aws.amazon.com/thread.jspa?threadID=162459
For end to end devops automation I want to have an environment on demand. For this I need to Spun up and environment on kubernetes which is eventually hosted on GCP.
My Use case
1. Developer Checks in the code in feature branch
2. Environment in Spun up on Google Cloud with Kubernetes
3. Application gets deployed on Kubernetes
4. Gets tested and then the environment gets destroyed.
I am able to do everything with Spinnaker except #2. i.e create Kube Cluster on GCP using Spinnaker.
Any help please
Thanks,
Amol
I'm not sure Spinnaker was meant for doing what the second point in your list. Spinnaker assumes a collection of resources (VM's or a Kubernetes cluster) and then works with that. So instead of spinning up a new GKE cluster Spinnaker makes use of existing clusters. I think it'd be better (for you costs as well ;) if you seperate the environments using Kubernetes namespaces.