10分钟快速开始SkyWalking结合Springboot项目
实习期间,公司让我去学习一下链路追踪如何集成到Springboot项目中。
为此有两个方案:
1.opentelementry+jaeger+prometheus
opentelementry 收集器收集线上的metrics和traces,然后发送给jaeger和prometheus去处理。jaeger确实好很多,在一个controller中用多线程调用另一个controller依然能显示一颗完整的调用过程。但是可视化做的不行,需要依赖于grafana。这就导致需要开很多服务,因此不太容易搭建。
2.SkyWalking+elasticsearch
skywalking集成了收集和分析的功能,所以只需要elasticsearch作为存储就行,简单,可视化好,适用于个人和中小公司使用。
依赖配置
xml
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
<version>2.2.0.RELEASE</version>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter</artifactId>
<version>2.2.0.RELEASE</version>
<exclusions>
<exclusion>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-logging</artifactId>
</exclusion>
</exclusions>
</dependency>
<dependency> <!-- 引入log4j2依赖 -->
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-log4j2</artifactId>
<version>2.2.0.RELEASE</version>
</dependency>
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-log4j-2.x</artifactId>
<version>9.1.0</version>
</dependency>
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-trace</artifactId>
<version>9.1.0</version>
</dependency>
这边需要排除掉springboot自带的日志框架,很重要
Dockerfile文件编写
yaml
version: '3.3'
services:
elasticsearch:
image: elasticsearch:7.17.6
container_name: elasticsearch
restart: always
ports:
- "9201:9200"
environment:
- "TAKE_FILE_OWNERSHIP=true" #volumes 挂载权限 如果不想要挂载es文件改配置可以删除
- "discovery.type=single-node" #单机模式启动
- "TZ=Asia/Shanghai" # 设置时区
- "ES_JAVA_OPTS=-Xms512m -Xmx512m" # 设置jvm内存大小
volumes:
- ./elasticsearch/logs:/usr/share/elasticsearch/logs
- ./elasticsearch/data:/usr/share/elasticsearch/data
#- ./elasticsearch/conf/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml
ulimits:
memlock:
soft: -1
hard: -1
skywalking-oap-server:
image: apache/skywalking-oap-server:8.9.1
container_name: skywalking-oap-server
depends_on:
- elasticsearch
links:
- elasticsearch
restart: always
ports:
- "11800:11800"
- "12800:12800"
environment:
SW_STORAGE: elasticsearch # 指定ES版本
SW_STORAGE_ES_CLUSTER_NODES: elasticsearch:9200
TZ: Asia/Shanghai
#volumes:
#- ./oap/conf/alarm-settings.yml:/skywalking/config/alarm-settings.yml
skywalking-ui:
image: apache/skywalking-ui:8.9.1
container_name: skywalking-ui
depends_on:
- skywalking-oap-server
links:
- skywalking-oap-server
restart: always
ports:
- "9090:8080"
environment:
SW_OAP_ADDRESS: http://skywalking-oap-server:12800
TZ: Asia/Shanghai
dockerfile如何运行,自行查询即可
启动完成之后,打开http://127.0.0.1:9090,会出现Skywalking的UI界面。
配置日志文件
在/src/main/resources下创建log4j2.xml文件
xml
<?xml version="1.0" encoding="UTF-8"?>
<Configuration status="INFO">
<Appenders>
<!-- 控制台输出 -->
<Console name="Console" target="SYSTEM_OUT">
<PatternLayout pattern="%d [%traceId] %-5p %c{1}:%L - %m%n"/>
</Console>
<!-- skywalking grpc 日志收集 8.4.0版本开始支持 -->
<GRPCLogClientAppender name="grpc-log">
<PatternLayout pattern="%d{HH:mm:ss.SSS} [%t] %-5level %logger{36} - %msg%n"/>
</GRPCLogClientAppender>
</Appenders>
<Loggers>
<Root level="INFO">
<AppenderRef ref="Console"/>
<AppenderRef ref="grpc-log"/>
</Root>
</Loggers>
</Configuration>
接下来,只需要写一个简单的测试项目,我这边主要用了我老的influxdb项目。可以参考一下,就不能cv大法了。
controller文件
java
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.*;
import top.warmheart.dao.DeviceDao;
import top.warmheart.pojo.Device;
import top.warmheart.service.impl.DeviceServiceImpl;
import top.warmheart.util.BaseResponse;
import java.time.Instant;
import java.time.LocalDateTime;
/**
* @author 滚~韬
* @date 2024/1/29 13:00
*/
@RestController
@RequestMapping("/influx")
public class InfluxDBController {
@Autowired
private DeviceServiceImpl deviceServiceImpl;
@Autowired
private DeviceDao dao;
@GetMapping("/queryByTime")
public BaseResponse Query(LocalDateTime start,LocalDateTime end){
return dao.QueryByTime(start,end);
}
@GetMapping("/queryById")
public BaseResponse Query(String Id){
return dao.QueryById(Id);
}
@PostMapping("/DeleteByTime")
public BaseResponse Delete(LocalDateTime start,LocalDateTime end){
return dao.DeleteByTime(start,end);
}
@PostMapping("/insertByBlocking")
public BaseResponse InsertByBlocking(Device device){
device.setTime(Instant.now());
return deviceServiceImpl.InsertDataByBlocking(device);
}
@PostMapping("/insert")
public BaseResponse Insert(Device device){
device.setTime(Instant.now());
return deviceServiceImpl.InsertData(device);
}
}
Service层
java
import com.influxdb.annotations.Measurement;
import com.influxdb.client.domain.InfluxQLQuery;
import com.influxdb.client.domain.WritePrecision;
import com.influxdb.exceptions.InfluxException;
import com.influxdb.query.InfluxQLQueryResult;
import com.influxdb.spring.influx.InfluxDB2Properties;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Service;
import top.warmheart.core.Op;
import top.warmheart.core.Query;
import top.warmheart.enums.ErrorCode;
import top.warmheart.model.DeleteModel;
import top.warmheart.model.QueryModel;
import top.warmheart.pojo.Device;
import top.warmheart.service.DeviceService;
import top.warmheart.util.BaseResponse;
import top.warmheart.util.InfluxdbUtils;
import java.time.OffsetDateTime;
import java.time.ZoneOffset;
import java.util.List;
import static top.warmheart.decorator.InfluxApiDecorator.*;
/**
* @author 滚~韬
* @date 2024/1/26 13:20
*/
@Service
public class DeviceServiceImpl implements DeviceService {
@Autowired
private InfluxDB2Properties influxDB2Properties;
protected static Logger log = LoggerFactory.getLogger(DeviceServiceImpl.class);
public BaseResponse QueryData(Class<?> clazz, QueryModel queryModel) {
Measurement annotation = clazz.getAnnotation(Measurement.class);
if (annotation != null) {
queryModel
.setMeasurement(annotation.name())
.setWhere(Op.where(queryModel));
}
String build = Query.build(queryModel);
return QueryData(build);
}
public BaseResponse QueryData(String sql) {
log.info("查询语句:" + sql);
InfluxQLQueryResult result = getInfluxQLQueryApi().query(new InfluxQLQuery(sql, influxDB2Properties.getBucket()));
return QueryData(result);
}
public BaseResponse QueryData(InfluxQLQueryResult result) {
if (result == null) {
return new BaseResponse(200, null, "获取成功,无数据");
}
List<Device> pojo = InfluxdbUtils.toPOJO(result, Device.class);
log.info("查询数据数量为:" + pojo.size() + "--------------------------");
return new BaseResponse(200, pojo, "获取成功");
}
public BaseResponse InsertData(Object o) {
try {
getWriteApi().writeMeasurement(WritePrecision.NS, o);
} catch (Exception e) {
return new BaseResponse(ErrorCode.SYSTEM_ERROR, "插入数据过程中异常");
}
return new BaseResponse(200, o, "插入成功");
}
public BaseResponse InsertDataByBlocking(Object o) {
try {
getWriteApiBlocking().writeMeasurement(WritePrecision.NS, o);
} catch (Exception e) {
return new BaseResponse(ErrorCode.SYSTEM_ERROR, "插入数据过程中异常");
}
return new BaseResponse(200, o, "插入成功");
}
/**
* 批量写有问题
*
* @param devices
* @return
*/
@Deprecated
public BaseResponse InsertData(List<Device> devices) {
try {
getWriteApi().writeMeasurements(WritePrecision.NS, devices);
} catch (Exception e) {
return new BaseResponse(ErrorCode.SYSTEM_ERROR, "插入数据过程中异常");
}
return new BaseResponse(200, devices, "插入成功");
}
public BaseResponse DeleteData(DeleteModel deleteModel) {
try {
OffsetDateTime startOff = OffsetDateTime.of(deleteModel.getStart(), ZoneOffset.UTC);
OffsetDateTime endOff = OffsetDateTime.of(deleteModel.getEnd(), ZoneOffset.UTC);
getDeleteApi().delete(startOff, endOff, "", influxDB2Properties.getBucket(), influxDB2Properties.getOrg());
} catch (InfluxException ie) {
log.warn("InfluxException: " + ie);
return new BaseResponse(ErrorCode.SYSTEM_ERROR, "删除错误");
}
return new BaseResponse(200, null, "删除成功");
}
}
dao层
java
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import top.warmheart.enums.ErrorCode;
import top.warmheart.model.DeleteModel;
import top.warmheart.model.QueryModel;
import top.warmheart.pojo.Device;
import top.warmheart.service.DeviceService;
import top.warmheart.util.BaseResponse;
import java.time.LocalDateTime;
import java.util.Map;
import java.util.TreeMap;
/**
* @Author:滚韬
* @Date:2024/1/30 14:28
*/
@Component
public class DeviceDao {
@Autowired
private DeviceService deviceService;
/**
* 根据给定时间范围查询数据
*
* @param start 开始时间
* @param end 结束时间,可选参数,如果不传,则默认为当前时间
* @return 查询结果的BaseResponse对象
*/
public BaseResponse QueryByTime(LocalDateTime start,LocalDateTime end){
QueryModel queryModel = new QueryModel();
if (start!=null){
queryModel.setStart(start);
if(end!=null){
queryModel.setEnd(end);
}else{
queryModel.setEnd(LocalDateTime.now());
}
}else {
return new BaseResponse(ErrorCode.SYSTEM_ERROR,"开始日期不能为空(检查是否格式正确)");
}
return deviceService.QueryData(Device.class, queryModel);
}
public BaseResponse QueryById(String Id){
Map<String, Object> map = new TreeMap<>();
map.put("device_no", Id);
QueryModel queryModel = new QueryModel();
queryModel.setMap(map);
return deviceService.QueryData(Device.class, queryModel);
}
public BaseResponse DeleteByTime(LocalDateTime start,LocalDateTime end){
DeleteModel deleteModel = new DeleteModel();
if (start!=null){
deleteModel.setStart(start);
if(end!=null){
deleteModel.setEnd(end);
}else{
deleteModel.setEnd(LocalDateTime.now());
}
}else {
return new BaseResponse(ErrorCode.SYSTEM_ERROR,"开始日期不能为空(检查是否格式正确)");
}
return deviceService.DeleteData(deleteModel);
}
}
启动
在虚拟机参数里加上这段,注意skywalking-agent.jar要去官网下载,jar包外面的文件也不能丢失,否则会报错
-javaagent:C:/skywalking/skywalking-agent/skywalking-agent.jar
-Dskywalking.agent.service_name=你自己的服务名字
-Dskywalking.collector.backend_service=127.0.0.1:11800
效果
请求查询接口,记得日志要打在service层里面
dashboard介绍
CPM/PPM:服务负荷
slow Services: 慢服务
Un-Health Services (Apdex): Apdex性能指标(1是满分)
Slow Endpoints: 慢请求点
Global Response Latency:百分比响应延时,不同百分比的延时时间,单位ms