Amazon EKS Connect to Kato
This article mainly explains how to install Kato on Amazon EKS and connect to Kato Cloud.
Before you begin, please check the following prerequisites:
Prepare an EKS Cluster
Before installing Kato, a standard kubernetes on EKS is required. We use AWS’s command line tool
eksctl to create an EKS cluster. For details, please refer to here. Select
1.17 for the cluster version,
t3.large (2 cores, 8GB) for node specifications, and
2 for computing nodes.
t3.large (2 cores, 8GB), 2 nodes is the minimum configuration. In a production environment, it is recommended to choose a higher configuration according to the actual situation.
- Kato’s minimum requirement for Kubernetes version is 1.13. It is recommended to select the highest version
1.17supported by EKS.
Connect to EKS Cluster via kubectl
After installing the EKS cluster, awsctl has set the configuration file for communicating with kube-apiserver to the local
~/.kube/config, we only need to install kubectl to communicate with the EKS cluster. For details, please refer to here.
Prepare an Elastic Load Balancing ELB
In order to ensure the high availability of the Kato gateway node (rbd-gateway), at least two gateway nodes need to be prepared, and before the traffic enters the Kato gateway node,
Add a 4-layer load balancing. We use AWS ELB as the 4-layer load balancer of the gateway node, and map the 5 ports of ELB
30008 to the corresponding ports of the gateway node . Please refer to Load Balancer ELB for the creation method.
Prepare an RDS
Prepare an RDS with a database engine of
MySQL and a version of
In addition, you need to create a database for Kato, the database name is
Install the Amazon FSx for Lustre CSI Driver
We use Amazon FSx to provide Kubernetes native file storage capabilities for Kato. We need to install a CSI driver for Amazon FSx. For details, please refer to here.
Linux users use the following command to install:
wget https://gridworkz-pkg.gitlab.com/pkg/helm && chmod +x helm && mv helm /usr/local/bin/
For MacOs and Windows users, please refer to Installing Helm.
Note: The version of helm needs to be 3.0 and above, helm 2.0+ is not supported.
Install Kato Operator
- Create namespace:
kubectl create ns rbd-system
- Install Kato Operator
# Add repository helm repo add kato https://openchart.gridworkz.com/gridworkz/kato # Installation helm install kato-operator-cloud kato/kato-operator \ --namespace rbd-system \ --version 1.1.0-cloud
To learn more about Kato Operator parameters, please refer to here.
- Confirm Kato Operator status
$ kubectl get pod -n rbd-system NAME READY STATUS RESTARTS AGE kato-operator-0 2/2 Running 0 110s
Wait for a while (depending on the specific network environment) until the status of kato-operator-0 becomes
Visit Kato Operator to Start the Installation of Kato.
Visit Kato Operator
Open the browser and enter the host IP address:
SERVER_ADDRESS is the address of the ELB prepared above
Installation Modeselect high availability
- Fill in the prepared RDS information into
Data Center Database
Build service running nodeSelect the two prepared computing nodes
Gateway public IPfill in the prepared ELB
- Shared storage: Select
Select the existing storage driver of the cluster->
- After completing the above configuration, click Configure ready to start the installation
When the installation progress is completed, it will jump to the following page:
The installation is complete.
Install Command Line Tools
wget https://gitlab.com/gridworkz/kato/releases/common/v5.2/grctl && chmod +x ./grctl mv ./grctl /usr/local/bin/grctl && /usr/local/bin/grctl install
wget https://pkg.kato.com/releases/common/darwin/v5.2/grctl && chmod +x ./grctl sudo mv ./grctl /usr/local/bin/grctl && /usr/local/bin/grctl install
If it outputs Install success, the installation is successful.
After the installation is complete, execute
grctl config to obtain
Region Config is used in the next step of
Connecting to Kato Cloud.
Connect with Kato Cloud
The specific docking method is as follows, add a cluster in the enterprise view cluster interface of the registered enterprise; the process is as follows:
Fill in the relevant information:
Cluster ID: Custom, cannot be modified after the cluster is added;
Cluster name: The name of the currently added cluster, customized, and can be modified after adding;
Remarks: The current cluster remarks information, just customize;
Region Config: The
Region Configfile defines the current cluster configuration information, including cluster
Web Socket communication address,
HTTP application default domain name suffix,
TCP application default access IP, etc. After adding the Region Config file when adding a cluster, the above information will be automatically read.
How to Use the Cluster
Select the cluster in Create Team and create a team belonging to the cluster. Going forward, the resources of the cluster will be used when creating components under the team.