Four Simple Ways To Deploy StarRocks

StarRocks, the high-performance real-time OLAP database, was designed to make data-intensive analytics fast and easy. When it comes to speed, it makes good on this promise by delivering query performance 5 to 10 times faster than other popular solutions. It can also perform real-time analytics while simultaneously updating historical records, and recent developments have enabled it to enhance real-time analytics with historical data from data lakes easily.

But what about ease of use? The StarRocks community values flexibility and low maintenance overhead, which is why you'll find it baked into every update. In this article, we'll take a look at one area StarRocks makes easy: deployment.

StarRocks provides many installation methods, but the best way to deploy StarRocks can depend on your specific needs and available technology. In the spirit of StarRocks' commitment to flexibility, we are going to explore four ways to deploy StarRocks and talk about how they compare to each other (with examples!). After reading this blog post, you will learn how different deployment methods compare and which is the best for you!

It is highly recommended that you get yourself familiar with StarRocks' architecture before diving into this article. Check out all the StarRocks documentation here. In the meantime, here's a quick overview.

StarRocks has a simple architecture with Frontend (FE) and Backend (BE) nodes:

• FEs are responsible for metadata management, client connection management, query planning, and query scheduling.

• BEs are responsible for data storage and SQL execution.

You can read more about them here.

Software used

• Docker

• MySQL (used as a client to connect to StarRocks)

Docker: Quick Functionality Test

The first and easiest way of deploying StarRocks is through a docker container.

To get docker, you can visit docker's official documentation here.

docker run -p 9030:9030 -p 8030:8030 -p 8040:8040 -itd starrocks/allin1-ubuntu

mysql -h127.0.0.1 -uroot -P9030

The Quick Start sets up the following:

• High Availability and distributed StarRocks

  • Front End node(s) for metadata management, client connections, etc.

  • Back End node(s) for data storage and SQL execution

• A Linux bastion host in an auto scaling group (optional)

• Supports deploying in an existing VPC or in a new VPC

Please refer to the quick start page for more details.

Deployment overview

From an AWS account, you will first launch Quick Start depending on whether you want to use an existing VPC or not.

Note: If you are using an existing VPC, please make sure you have the S3 endpoint configured. See the documentation here.

An AWS CloudFormation console will pre-populate the template with some common options, but feel free to change as needed.

Here we highlight some of the configurations that might be confusing.

You will be able to review and confirm the template settings before choosing “Create stack” to deploy the stack.

When the status is CREATE_COMPLETE, the Starrocks deployment is ready.

Please review the detailed deployment steps on AWS for more help launching QuickStart. It should take only a few minutes to review, and between 10 to 30 minutes to kick it off.

StarRocks Kubernetes Operator: Fully fledged StarRocks on K8s

The StarRocks Kubernetes Operator is a useful tool developed by the StarRocks community. It can help you deploy StarRocks' Front End (FE), Back End (BE), and Compute Node (CN) components into your Kubernetes (K8s) environment with a few simple commands. This deployment method supports deploying any number of FE (at least 1 for FE), BE, and CN nodes in 2 simple steps.

Prerequisites

This guide assumes that you already have a functioning Kubernetes environment at hand. If you do not have one, there are several options that you can choose from:

Quick validation test:

• Docker Desktop with Kubernetes

• Minikube

Managed K8s on the public cloud:

• EKS from AWS

• GKE from GCP

Step 1. Install StarRocks' K8s operator

The first step is to install the starrockscluster CRD:

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

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

kubectl apply -f custom-namespace-operator.yaml

wget https://raw.githubusercontent.com/StarRocks/starrocks-kubernetes-operator/main/examples/starrocks/starrocks-fe-and-be.yaml

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

kubectl get svc -n starrocks

mysql -h <FE-IP> -P9030 -urooot

kubectl -n starrocks edit src starrockscluster-sample

starRocksFeSpec:
image: starrocks/fe-ubuntu:x.x.x
replicas: 3
requests:
cpu: 4
memory: 16Gi
service: 
type: LoadBalancer ## change to LoadBalancer

kubectl get svc -n starrocks

mysql -h <FE-IP> -P9030 -urooot