CelerData-Kubernetes-Operator

English | 中文

Overview

CelerData Kubernetes Operator is a project that implements the deployment and operation of CelerData, a next-generation sub-second MPP OLAP database, on Kubernetes. It facilitates the deployment of CelerData' Frontend (FE), Backend (BE), and Compute Node (CN) components within your Kubernetes environment. It also includes Helm chart for easy installation and configuration. With CelerData Kubernetes Operator, you can easily manage the lifecycle of CelerData clusters, such as installing, scaling, upgrading etc.

Note

The CelerData k8s operator was designed to be a level 2 operator. See https://sdk.operatorframework.io/docs/overview/operator-capabilities/ to understand more about the capabilities of a level 2 operator.

Prerequisites

1. Kubernetes version >= 1.23.0
2. Helm version >= 3.0

Features

Operator Features

• Support deploying CelerData FE, BE and CN components separately FE component is a must-have component, BE and CN components can be optionally deployed
• Support multiple CelerData clusters in one Kubernetes cluster
• Support external clients outside the network of kubernetes to load data into CelerData using STREAM LOAD
• Support automatic scaling for CN nodes based on CPU and memory usage
• Support mounting persistent volumes for CelerData containers

Helm Chart Features

• Support Helm Chart for easy installation and configuration
  • using kube-celerdata Helm chart to install both operator and CelerData cluster
  • using operator Helm Chart to install operator, and using CelerData Helm Chart to install celerdata cluster
• Support initializing the password of root in your CelerData cluster during installation.
• Support integration with other components in the Kubernetes ecosystem, such as Prometheus, Datadog, etc.

Installation

In order to use CelerData in Kubernetes, you need to install:

1. CelerDataCluster CRD
2. CelerData Operator
3. CelerDataCluster CR

There are two ways to install Operator and CelerData Cluster.

1. Install Operator and CelerData Cluster by yaml Manifest.
2. Install Operator and CelerData Cluster by Helm Chart.

Note: In every release, we will provide the latest version of the yaml Manifest and Helm Chart. You can find them in https://github.com/celerdata/celerdata-kubernetes-operator/releases

Installation by yaml Manifest

Please see Deploy CelerData With Operator document for more details.

1. Apply the CelerDataCluster CRD

kubectl apply -f https://raw.githubusercontent.com/celerdata/celerdata-kubernetes-operator/main/deploy/celerdata.com_celerdataclusters.yaml

2. Apply the Operator manifest

Apply the Operator manifest. By default, the Operator is configured to install in the CelerData namespace. To use the Operator in a custom namespace, download the Operator manifest and edit all instances of namespace: CelerData to specify your custom namespace. Then apply this version of the manifest to the cluster with kubectl apply -f {local-file-path} instead of using the command below.

kubectl apply -f https://raw.githubusercontent.com/celerdata/celerdata-kubernetes-operator/main/deploy/operator.yaml

3. Deploy the CelerData cluster

You need to prepare a separate yaml file to deploy the CelerData. The celerdata cluster CRD fields explains in api.md. The examples directory contains some simple example for reference.

You can use any of the template yaml file as a starting point. You can further add more configurations into the template yaml file following this deployment documentation.

For demonstration purpose, we use the celerdata-fe-and-be.yaml example template to start a 3 FE and 3 BE CelerData cluster.

Here's an example yaml for Docker Desktop with local desktop access with CelerData 3.2.1 so you can upgrade in later steps.

atwong@Albert-CelerData celerdatatest % cat celerdata-fe-and-be.yaml
apiVersion: celerdata.com/v1
kind: CelerDataCluster
metadata:
  name: celerdatacluster-sample
  namespace: celerdata
spec:
  celerDataFeSpec:
    image: us-west1-docker.pkg.dev/phrasal-verve-350013/celerdata/fe-ubuntu:3.2.1
    replicas: 3
    requests:
      cpu: 1
      memory: 2Gi
    limits:
      cpu: 4
      memory: 16Gi
    service:            
      type: LoadBalancer
  celerDataBeSpec:
    image: us-west1-docker.pkg.dev/phrasal-verve-350013/celerdata/be-ubuntu:3.2.1
    replicas: 3
    requests:
      cpu: 1
      memory: 2Gi
    limits:
      cpu: 4
      memory: 8Gi

kubectl apply -f celerdata-fe-and-be.yaml

4. Connect the CelerData cluster

To connect, just use the mysql client and connect to the CelerData cluster port 9030. An example of a connection is shown below.

Note

If you want to connect remotely or through your desktop, you will need to enable the k8s Load Balander.

kubectl -n celerdata get svc

NAME                                 TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)                                                       AGE
celerdatacluster-sample-be-search    ClusterIP      None            <none>        9050/TCP                                                      5m2s
celerdatacluster-sample-be-service   ClusterIP      10.103.248.52   <none>        9060/TCP,8040/TCP,9050/TCP,8060/TCP                           5m2s
celerdatacluster-sample-fe-search    ClusterIP      None            <none>        9030/TCP                                                      6m22s
celerdatacluster-sample-fe-service   LoadBalancer   10.99.14.222    localhost     8030:32326/TCP,9020:32578/TCP,9030:30774/TCP,9010:32505/TCP   6m22s

mysql -h 127.0.0.1 -P 9030 -uroot

Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 3
Server version: 5.1.0 3.2.1-79ee91d

Copyright (c) 2000, 2024, Oracle and/or its affiliates.

Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.

Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.

mysql>

5. Upgrade the CelerData cluster

To upgrade, just patch the CelerData cluster.

kubectl -n celerdata patch celerdatacluster celerdatacluster-sample --type='merge' -p '{"spec":{"celerDataFeSpec":{"image":"us-west1-docker.pkg.dev/phrasal-verve-350013/celerdata/fe-ubuntu:latest"}}}'
kubectl -n celerdata patch celerdatacluster celerdatacluster-sample --type='merge' -p '{"spec":{"celerDataBeSpec":{"image":"us-west1-docker.pkg.dev/phrasal-verve-350013/celerdata/be-ubuntu:latest"}}}'

6. Resize the CelerData cluster

To resize, just patch the CelerData cluster.

Important

Once you deploy with 3 FE nodes, you are in HA mode. Do not resize FE nodes below 3 since that will affect cluster quorum. This rule doesn't apply to CN nodes.

kubectl -n celerdata patch celerdatacluster celerdatacluster-sample --type='merge' -p '{"spec":{"celerDataBeSpec":{"replicas":9}}}'

7. Delete/stop the CelerData cluster

To delete/stop the CelerData cluster, just execute the delete command.

kubectl delete -f celerdata-fe-and-be.yaml

or

kubectl delete celerdatacluster celerdatacluster-sample -n celerdata

8. Delete/stop the CelerData Operator

To delete/stop the CelerData Operate, just execute the delete command.

kubectl delete -f https://raw.githubusercontent.com/celerdata/celerdata-kubernetes-operator/main/deploy/operator.yaml

Installation by Helm Chart

Please see kube-celerdata for how to install both operator and CelerData cluster by Helm Chart.

If you want more flexibility in managing your CelerData clusters, you can deploy Operator using operator Helm Chart and CelerData using celerdata Helm Chart separately.

Other Documents

• In doc directory, you can find more documents about how to use CelerData Operator.
• In examples directory, you can find more examples about how to write CelerDataCluster CR.