Can a cloud native architecture improve elasticity and stability in the context of a simulated Electronic Health Record(EHR) system?
Run the SQL FileDatabase/data.sqlin MySQL Server
change the application.properties file in /Prototype/src/main/resources
#Service configuration
##service name
spring.application.name=backend_healthcare
##service port
server.port=6061
#Database configuration
##MySQL database configuration
spring.datasource.url=jdbc:mysql:https://localhost:3306/{DatabaseName}?useUnicode=true&characterEncoding=utf8&serverTimezone=GMT%2B8
spring.datasource.username=root
spring.datasource.password=6021
spring.datasource.driver-class-name=com.mysql.cj.jdbc.Driver
##(Database framework)Mybatis log configuration
#mybatis-plus.configuration.log-impl=org.apache.ibatis.logging.stdout.StdOutImplUse Maven to clean the dependencies and package the project
Get the jar file in /Prototype/target/
Use command below in any Server which have Java:17 environment
java -jar app.jarAnd the Running logs will be show in the terminal
The project uses four CentOS 8.2 cloud servers. One server is used to deploy cloud native components Nacos, Sentinel, and Nginx. The other three servers are used to deploy microservices.
Install docker dependencies
#centos
yum install -y yum-utils device-mapper-persistent-data lvm2
yum install docker-ceEnable start docker service automatically
systemctl enable docker.serviceCheck the status of docker.service
systemctl status docker.servicedocker pull mysql image
docker pull mysql:8.0.28run MySQL command
docker run -p 3306:3306 --name mysql --restart=unless-stopped --privileged=true \
-v /opt/docker/mysql8.0.28/mysql:/etc/mysql \
-v /opt/docker/mysql8.0.28/logs:/logs \
-v /opt/docker/mysql8.0.28/data:/var/lib/mysql \
-e MYSQL_ROOT_PASSWORD=6021 -d mysql:8.0.28check running status and logs
docker ps
docker logs `name`Pull Nacos image
docker pull nacos/nacos-server:v2.1.1Run command
docker run -d -p 8848:8848 -p 9848:9848 --name nacos --restart=unless-stopped \
-e jvm_xms=256m \
-e jvm_xmx=256m \
-e MODE=standalone \
nacos/nacos-server:v2.1.1pull Sentinel image
docker pull bladex/sentinel-dashboardrun Sentinel
docker run --name sentinel --restart=unless-stopped -d -p 8858:8858 -p 8719:8719 -d bladex/sentinel-dashboardPull Official Nginx Image
docker pull nginx:latestUse DockerFile to build image
#/-- DockerFile
# /-- Nginx
# /-- DockerFile
#use the official Nginx image as the base image
FROM nginx:latest
#Copy local nginx configuration file to the container
COPY nginx.conf /etc/nginx/nginx.conf
# Expose ports.
EXPOSE 80
EXPOSE 8081
EXPOSE 7071
EXPOSE 9091
# Define default command.
CMD ["nginx", "-g", "daemon off;"]Update nginx.conf, it s the Nginx configuration file
worker_processes 1;
events {
worker_connections 1024;
}
http {
upstream backend_servers_8081 {
server 192.168.1.101:8081;
server 192.168.1.100:8081;
server 192.168.9.67:8081;
}
upstream backend_servers_9091 {
server 192.168.1.101:9091;
server 192.168.1.100:9091;
server 192.168.9.67:9091;
}
upstream backend_servers_7071 {
server 192.168.1.101:7071;
server 192.168.1.100:7071;
server 192.168.9.67:7071;
}
server {
listen 80;
listen 8081;
server_name localhost;
location / {
proxy_pass https://backend_servers_8081;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
}
}
server {
listen 7071;
server_name localhost;
location / {
proxy_pass https://backend_servers_7071;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
}
}
server {
listen 9091;
server_name localhost;
location / {
proxy_pass https://backend_servers_9091;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
}
}
}Build My Nginx Image
/--Nginx
/--nginx.conf
/--DockerFile
#bulid command
docker build -t nginx_ghc .
#check docker images
docker ps
#run commmand
docker run -d --name docker_ghc -p 80:80 -p 9091:9091 -p 7071:7071 -p 8081:8081 nginx
#check running status
Before this stage, check the docker running status by command docker ps , make sure that all component run successfully
Update the application.properties file
/--Cloud_Native_Architecture
/--appointmentservice1
/--application.properties server.port=9091
/--patientservice1
/--application.properties server.port=7071
/--userservice1
/--application.properties server.port=8081and sentinel, nacos, mysql configuration
#Sentinel configuration
##port for uploading runtime information to sentinel dashboard
spring.cloud.sentinel.transport.port=8719
##remote address of sentinel dashboard
spring.cloud.sentinel.transport.dashboard=192.168.0.100:8858
#Nacos configuration
##remote address of nacos server
spring.cloud.nacos.discovery.server-addr=192.168.0.100:8848
spring.cloud.loadbalancer.nacos.enabled=true
#Database configuration
##MySQL database configuration
spring.datasource.url=jdbc:mysql:https://192.168.0.100:3306/health_g?useUnicode=true&characterEncoding=utf8&serverTimezone=GMT%2B8
spring.datasource.username=root
spring.datasource.password=6021
spring.datasource.driver-class-name=com.mysql.cj.jdbc.Driver
###(Database framework)Mybatis log configuration
#mybatis-plus.configuration.log-impl=org.apache.ibatis.logging.stdout.StdOutImplMaven run clean and package to get userService.jar , patientService.jar and appointmentService.jar
Upload it in the Linux Server and use Docker to build container by DockerFile
#java version is 17
FROM openjdk:17
#copy all jar files
COPY *.jar /app.jar
#server port
CMD ["--server.port=9091"]
#container expose port
EXPOSE 9091
#container run command
ENTRYPOINT ["java","-jar","/app.jar"]DockerFile is in Cloud_Native_Architecture/userservice1/DockerFile Cloud_Native_Architecture/patientservice1/DockerFile Cloud_Native_Architecture/appointmentservice1/DockerFile
#place the jar file and DockerFile in same Folders
/--patientservice
/--patientService.jar
/--DockerFile
/--userservice
/--userService.jar
/--DockerFile
/--appointmentservice
/--appointmentService.jar
/--Dockerfile
#Build All Service to image
docker build -t patientservice /root/patientservice && \
docker build -t appointmentservice /root/appointmentservice && \
docker build -t userservice /root/userservice
#show all exist images
docker images
Run Command
#docker command
docker run -d -p 8081:8081 --network=host --restart=unless-stopped --name userservice
docker run -d -p 7071:7071 --network=host --restart=unless-stopped --name patientservice
docker run -d -p 9091:9091 --network=host --restart=unless-stopped --name appointmentservice
#show all running container
docker ps
And run it in another server
Use adrianhajdin/healthcare project to build the frontend, simulating real-world EHR system scenarios
cd g_healthcare_ar/FrontEnd
npm install
npm run devUse CodeArts TestPlan to do the performance testing. And compare cloud native architecture system with prototype
Prototype: HTTP Post request, request URL: “https://123.249.78.235:6061/user/”
Cloud-Native Architecture: HTTP Post request, request URL: “https://114.116.231.193/user”
Request Body:{
"userId": "$UUID",
"email": "[email protected]",
"phone": "+&random number in 1-50 $UUID",
"name": "$UUID"
}
Starting Concurrency: 5
Total Concurrency Increment: 400
Duration per Stage: 1 minute
Increment Rate: 20%
Performance: The Cloud Native Architecture clearly outperforms in all key performance metrics. It has a higher success rate (100% vs. 98.71%), shorter average response time, and lower maximum response time. Additionally, there were no instances of connection refusal or response timeout.
Stability: The Cloud Native Architecture shows significantly better stability compared to the Prototype Architecture. This is evident from its 0 failed requests and no warning states. The maximum response times at various TP levels are all lower than those of the Prototype, especially at TP99, TP99.9, and TP99.99, where the response time differences are significant. This indicates that even under high load and high concurrency, the Cloud Native Architecture maintains stable performance.
Conclusion: Based on these comparative results, it can be concluded that the Cloud Native Architecture has clear advantages in terms of performance, stability, and resource utilization. It provides better support and lower response times in high concurrency and high load environments. These advantages make the Cloud Native Architecture more suitable for modern applications, especially in scenarios requiring high load, high scalability, and high reliability.
Detail in [Cloud Native Architecture Performance Testing Report.pdf](Performance Testing Report\Cloud Native Architecture Performance Testing Report.pdf) and [Prototype Performance Testing Report.pdf](Performance Testing Report\Prototype Performance Testing Report.pdf)
This is the script for running and building the application which can be used as a reference
/DockerFile/container_run_shell
+---/DockerFile/container_run_shell
| appointmentservice.sh
| build all service image.sh
| patientservice.sh
| stop & delete & remove all service image.sh
| userservice.sh#build all service
docker build -t patientservice /root/patientservice && \
docker build -t appointmentservice /root/appointmentservice && \
docker build -t userservice /root/userservice
############################################################################################################
#run appointmentservice
PREFIX="appointmentservice"
SUFFIX=$(date +%Y%m%d%H%M%S)
NAME="$PREFIX-$SUFFIX"
docker run -d -p 9091:9091 --network=host --restart=unless-stopped --name "$NAME" appointmentservice
############################################################################################################
#run userservice
PREFIX="userservice"
SUFFIX=$(date +%Y%m%d%H%M%S)
NAME="$PREFIX-$SUFFIX"
docker run -d -p 8081:8081 --network=host --restart=unless-stopped --name "$NAME" userservice
############################################################################################################
#run appointmentservice
PREFIX="patientservice"
SUFFIX=$(date +%Y%m%d%H%M%S)
NAME="$PREFIX-$SUFFIX"
docker run -d -p 7071:7071 --network=host --restart=unless-stopped --name "$NAME" patientservice
############################################################################################################
#stop & remove all service and images
docker stop $(docker ps -q) && docker rm $(docker ps -aq) && docker rmi $(docker images -q --filter=reference='*service*')