- Sentinel规则推送模式
Sentinel规则的推送有下面三种模式:
1.1 原始模式
如果不做任何修改,Dashboard 的推送规则方式是通过 API 将规则推送至客户端并直接更新到内存中:
这种做法的好处是简单,无依赖;坏处是应用重启规则就会消失,仅用于简单测试,不能用于生产环境。
1.2 拉模式
pull 模式的数据源(如本地文件、RDBMS 等)一般是可写入的。使用时需要在客户端注册数据源:将对应的读数据源注册至对应的 RuleManager,将写数据源注册至 transport 的 WritableDataSourceRegistry 中。
首先 Sentinel 控制台通过 API 将规则推送至客户端并更新到内存中,接着注册的写数据源会将新的规则保存到本地的文件中。使用 pull 模式的数据源时一般不需要对 Sentinel 控制台进行改造。这种实现方法好处是简单,坏处是无法保证监控数据的一致性。
官方demo: sentinel-demo/sentinel-demo-dynamic-file-rule
引入依赖
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-datasource-extension</artifactId>
<version>1.8.0</version>
</dependency>
核心代码:
// FileRefreshableDataSource 会周期性的读取文件以获取规则,当文件有更新时会及时发现,并将规则更新到内存中。
ReadableDataSource<String, List<FlowRule>> ds = new FileRefreshableDataSource<>(
flowRulePath, source -> JSON.parseObject(source, new TypeReference<List<FlowRule>>() {})
);
// 将可读数据源注册至 FlowRuleManager.
FlowRuleManager.register2Property(ds.getProperty());
WritableDataSource<List<FlowRule>> wds = new FileWritableDataSource<>(flowRulePath, this::encodeJson);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
WritableDataSourceRegistry.registerFlowDataSource(wds);
拉模式改造
实现InitFunc接口,在init中处理DataSource初始化逻辑,并利用spi机制实现加载。
其中部分核心代码
public class FileDataSourceInit implements InitFunc {
@Override
public void init() throws Exception {
//创建文件存储目录
RuleFileUtils.mkdirIfNotExits(PersistenceRuleConstant.storePath);
//创建规则文件
RuleFileUtils.createFileIfNotExits(PersistenceRuleConstant.rulesMap);
//处理流控规则逻辑
dealFlowRules();
// 处理降级规则
dealDegradeRules();
// 处理系统规则
dealSystemRules();
// 处理热点参数规则
dealParamFlowRules();
// 处理授权规则
dealAuthRules();
}
private void dealFlowRules() throws FileNotFoundException {
String ruleFilePath = PersistenceRuleConstant.rulesMap.get(PersistenceRuleConstant.FLOW_RULE_PATH).toString();
//创建流控规则的可读数据源
ReadableDataSource<String, List<FlowRule>> flowRuleRDS = new FileRefreshableDataSource(
ruleFilePath, RuleListConverterUtils.flowRuleListParser
);
// 将可读数据源注册至FlowRuleManager 这样当规则文件发生变化时,就会更新规则到内存
FlowRuleManager.register2Property(flowRuleRDS.getProperty());
WritableDataSource<List<FlowRule>> flowRuleWDS = new FileWritableDataSource<List<FlowRule>>(
ruleFilePath, RuleListConverterUtils.flowFuleEnCoding
);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
WritableDataSourceRegistry.registerFlowDataSource(flowRuleWDS);
}
private void dealDegradeRules() throws FileNotFoundException {
//讲解规则文件路径
String degradeRuleFilePath = PersistenceRuleConstant.rulesMap.get(PersistenceRuleConstant.DEGRAGE_RULE_PATH).toString();
//创建流控规则的可读数据源
ReadableDataSource<String, List<DegradeRule>> degradeRuleRDS = new FileRefreshableDataSource(
degradeRuleFilePath, RuleListConverterUtils.degradeRuleListParse
);
// 将可读数据源注册至FlowRuleManager 这样当规则文件发生变化时,就会更新规则到内存
DegradeRuleManager.register2Property(degradeRuleRDS.getProperty());
WritableDataSource<List<DegradeRule>> degradeRuleWDS = new FileWritableDataSource<>(
degradeRuleFilePath, RuleListConverterUtils.degradeRuleEnCoding
);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
WritableDataSourceRegistry.registerDegradeDataSource(degradeRuleWDS);
}
private void dealSystemRules() throws FileNotFoundException {
//讲解规则文件路径
String systemRuleFilePath = PersistenceRuleConstant.rulesMap.get(PersistenceRuleConstant.SYSTEM_RULE_PATH).toString();
//创建流控规则的可读数据源
ReadableDataSource<String, List<SystemRule>> systemRuleRDS = new FileRefreshableDataSource(
systemRuleFilePath, RuleListConverterUtils.sysRuleListParse
);
// 将可读数据源注册至FlowRuleManager 这样当规则文件发生变化时,就会更新规则到内存
SystemRuleManager.register2Property(systemRuleRDS.getProperty());
WritableDataSource<List<SystemRule>> systemRuleWDS = new FileWritableDataSource<>(
systemRuleFilePath, RuleListConverterUtils.sysRuleEnCoding
);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
WritableDataSourceRegistry.registerSystemDataSource(systemRuleWDS);
}
private void dealParamFlowRules() throws FileNotFoundException {
//讲解规则文件路径
String paramFlowRuleFilePath = PersistenceRuleConstant.rulesMap.get(PersistenceRuleConstant.HOT_PARAM_RULE).toString();
//创建流控规则的可读数据源
ReadableDataSource<String, List<ParamFlowRule>> paramFlowRuleRDS = new FileRefreshableDataSource(
paramFlowRuleFilePath, RuleListConverterUtils.paramFlowRuleListParse
);
// 将可读数据源注册至FlowRuleManager 这样当规则文件发生变化时,就会更新规则到内存
ParamFlowRuleManager.register2Property(paramFlowRuleRDS.getProperty());
WritableDataSource<List<ParamFlowRule>> paramFlowRuleWDS = new FileWritableDataSource<>(
paramFlowRuleFilePath, RuleListConverterUtils.paramRuleEnCoding
);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
ModifyParamFlowRulesCommandHandler.setWritableDataSource(paramFlowRuleWDS);
}
private void dealAuthRules() throws FileNotFoundException {
//讲解规则文件路径
String authFilePath = PersistenceRuleConstant.rulesMap.get(PersistenceRuleConstant.AUTH_RULE_PATH).toString();
//创建流控规则的可读数据源
ReadableDataSource<String, List<AuthorityRule>> authRuleRDS = new FileRefreshableDataSource(
authFilePath, RuleListConverterUtils.authorityRuleParse
);
// 将可读数据源注册至FlowRuleManager 这样当规则文件发生变化时,就会更新规则到内存
AuthorityRuleManager.register2Property(authRuleRDS.getProperty());
WritableDataSource<List<AuthorityRule>> authRuleWDS = new FileWritableDataSource<>(
authFilePath, RuleListConverterUtils.authorityEncoding
);
// 将可写数据源注册至 transport 模块的 WritableDataSourceRegistry 中.
// 这样收到控制台推送的规则时,Sentinel 会先更新到内存,然后将规则写入到文件中.
WritableDataSourceRegistry.registerAuthorityDataSource(authRuleWDS);
}
}
1.3 推模式
生产环境下一般更常用的是 push 模式的数据源。对于 push 模式的数据源,如远程配置中心(ZooKeeper, Nacos, Apollo等等),推送的操作不应由 Sentinel 客户端进行,而应该经控制台统一进行管理,直接进行推送,数据源仅负责获取配置中心推送的配置并更新到本地。因此推送规则正确做法应该是 配置中心控制台/Sentinel 控制台 → 配置中心 → Sentinel 数据源 → Sentinel,而不是经 Sentinel 数据源推送至配置中心。这样的流程就非常清晰了:
1.3.1 基于Nacos配置中心控制台实现推送
官方demo: sentinel-demo-nacos-datasource
引入依赖
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-datasource-nacos</artifactId>
<version>1.8.0</version>
</dependency>
核心代码
// nacos server ip
private static final String remoteAddress = "localhost:8848";
// nacos group
private static final String groupId = "Sentinel:Demo";
// nacos dataId
private static final String dataId = "com.alibaba.csp.sentinel.demo.flow.rule";
ReadableDataSource<String, List<FlowRule>> flowRuleDataSource = new NacosDataSource<>(remoteAddress, groupId, dataId,source -> JSON.parseObject(source, new TypeReference<List<FlowRule>>() {}));
FlowRuleManager.register2Property(flowRuleDataSource.getProperty());
nacos配置中心中配置流控规则
[
{
"resource": "TestResource",
"controlBehavior": 0,
"count": 10.0,
"grade": 1,
"limitApp": "default",
"strategy": 0
}
]
微服务中通过yml配置实现
SentinelProperties 内部提供了 TreeMap 类型的 datasource 属性用于配置数据源信息
1)引入依赖
<!--sentinel持久化 采用 Nacos 作为规则配置数据源-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-datasource-nacos</artifactId>
</dependency>
2)yml中配置
spring:
application:
name: mall-user-sentinel-demo
cloud:
nacos:
discovery:
server-addr: 127.0.0.1:8848
sentinel:
transport:
# 添加sentinel的控制台地址
dashboard: 127.0.0.1:8080
# 指定应用与Sentinel控制台交互的端口,应用本地会起一个该端口占用的HttpServer
port: 8719
datasource:
ds1:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}
groupId: DEFAULT_GROUP
data-type: json
rule-type: flow
源码参考com.alibaba.cloud.sentinel.datasource.config.AbstractDataSourceProperties#postRegister
3)nacos配置中心中添加
[
{
"resource": "userinfo",
"limitApp": "default",
"grade": 1,
"count": 1,
"strategy": 0,
"controlBehavior": 0,
"clusterMode": false
}
]
缺点:直接在Sentinel Dashboard中修改规则配置,配置中心的配置不会发生变化
思考: 如何实现将通过sentinel控制台设置的规则直接持久化到nacos配置中心?
扩展改造的思路:
Sentinel Dashboard监听Nacos配置的变化,如发生变化就更新本地缓存。在Sentinel Dashboard端新增或修改规则配置在保存到内存的同时,直接发布配置到nacos配置中心;Sentinel Dashboard直接从nacos拉取所有的规则配置。
sentinel Dashboard和sentinel client 不直接通信,而是通过nacos配置中心获取到配置的变更。
1.3.2 基于Sentinel控制台实现推送
从 Sentinel 1.4.0 开始,Sentinel 控制台提供 DynamicRulePublisher 和 DynamicRuleProvider 接口用于实现应用维度的规则推送和拉取:
- DynamicRuleProvider: 拉取规则
- DynamicRulePublisher: 推送规则
Sentinel Dashboard改造
第1步:在com.alibaba.csp.sentinel.dashboard.rule包下创建nacos包,然后把各种场景的配置规则拉取和推送的实现类写到此包下
可以参考Sentinel Dashboard test包下的流控规则拉取和推送的实现逻辑:
注意:微服务接入Sentinel client,yml配置需要匹配对应的规则后缀
第2步:进入com.alibaba.csp.sentinel.dashboard.controller包下修改对应的规则controller实现类
以流控规则为例,从Nacos配置中心获取所有的流控规则
@GetMapping("/rules")
@AuthAction(PrivilegeType.READ_RULE)
public Result<List<FlowRuleEntity>> apiQueryMachineRules(@RequestParam String app,
@RequestParam String ip,
@RequestParam Integer port) {
if (StringUtil.isEmpty(app)) {
return Result.ofFail(-1, "app can't be null or empty");
}
if (StringUtil.isEmpty(ip)) {
return Result.ofFail(-1, "ip can't be null or empty");
}
if (port == null) {
return Result.ofFail(-1, "port can't be null");
}
try {
// List<FlowRuleEntity> rules = sentinelApiClient.fetchFlowRuleOfMachine(app, ip, port);
//从配置中心获取规则配置
List<FlowRuleEntity> rules = ruleProvider.getRules(app, ip, port);
rules = repository.saveAll(rules);
return Result.ofSuccess(rules);
} catch (Throwable throwable) {
logger.error("Error when querying flow rules", throwable);
return Result.ofThrowable(-1, throwable);
}
}
新增流控规则,会推送到nacos配置中心
@PostMapping("/rule")
@AuthAction(PrivilegeType.WRITE_RULE)
public Result<FlowRuleEntity> apiAddFlowRule(@RequestBody FlowRuleEntity entity) {
Result<FlowRuleEntity> checkResult = checkEntityInternal(entity);
if (checkResult != null) {
return checkResult;
}
entity.setId(null);
Date date = new Date();
entity.setGmtCreate(date);
entity.setGmtModified(date);
entity.setLimitApp(entity.getLimitApp().trim());
entity.setResource(entity.getResource().trim());
try {
entity = repository.save(entity);
//publishRules(entity.getApp(), entity.getIp(), entity.getPort()).get(5000, TimeUnit.MILLISECONDS);
//发布规则到配置中心
publishRules(entity.getApp());
return Result.ofSuccess(entity);
} catch (Throwable t) {
Throwable e = t instanceof ExecutionException ? t.getCause() : t;
logger.error("Failed to add new flow rule, app={}, ip={}", entity.getApp(), entity.getIp(), e);
return Result.ofFail(-1, e.getMessage());
}
}
/**
* 发布到配置中心
* @param app
* @throws Exception
*/
private void publishRules(/*@NonNull*/ String app) throws Exception {
List<FlowRuleEntity> rules = repository.findAllByApp(app);
rulePublisher.publish(app, rules);
}
测试:微服务接入改造后的Sentinel Dashboard
引入依赖
<!--sentinel持久化 采用 Nacos 作为规则配置数据源-->
<dependency>
<groupId>com.alibaba.csp</groupId>
<artifactId>sentinel-datasource-nacos</artifactId>
</dependency>
增加yml配置
server:
port: 8806
spring:
application:
name: mall-user-sentinel-rule-push-demo #微服务名称
#配置nacos注册中心地址
cloud:
nacos:
discovery:
server-addr: 127.0.0.1:8848
sentinel:
transport:
# 添加sentinel的控制台地址
dashboard: 127.0.0.1:8080
# 指定应用与Sentinel控制台交互的端口,应用本地会起一个该端口占用的HttpServer
#port: 8719
datasource:
# ds1: #名称自定义,唯一
# nacos:
# server-addr: 127.0.0.1:8848
# dataId: ${spring.application.name}
# groupId: DEFAULT_GROUP
# data-type: json
# rule-type: flow
flow-rules:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}-flow-rules
groupId: SENTINEL_GROUP # 注意groupId对应Sentinel Dashboard中的定义
data-type: json
rule-type: flow
degrade-rules:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}-degrade-rules
groupId: SENTINEL_GROUP
data-type: json
rule-type: degrade
param-flow-rules:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}-param-flow-rules
groupId: SENTINEL_GROUP
data-type: json
rule-type: param-flow
authority-rules:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}-authority-rules
groupId: SENTINEL_GROUP
data-type: json
rule-type: authority
system-rules:
nacos:
server-addr: 127.0.0.1:8848
dataId: ${spring.application.name}-system-rules
groupId: SENTINEL_GROUP
data-type: json
rule-type: system
以流控规则测试,当在sentinel dashboard配置了流控规则,会在nacos配置中心生成对应的配置。
热点参数规则失效和解决思路
注意:控制台改造后有可能出现规则不生效的情况,比如热点参数规则因为Converter解析json错误的原因会导致不生效。
参见源码:com.alibaba.csp.sentinel.datasource.AbstractDataSource#loadConfig(S) 会解析配置规则。
原因是:改造dashboard,提交到nacos配置中心的数据是ParamFlowRuleEntity类型,微服务拉取配置要解析的是ParamFlowRule类型,会导致规则解析丢失数据,造成热点规则不生效。 其他的规则原理也是一样,存在失效的风险。
nacos配置中心保存的数据格式:
[{
"app": "mall-user-sentinel-rule-push-demo",
"gmtCreate": 1616136838785,
"gmtModified": 1616136838785,
"id": 1,
"ip": "192.168.3.1",
"port": 8719,
"rule": {
"burstCount": 0,
"clusterConfig": {
"fallbackToLocalWhenFail": true,
"sampleCount": 10,
"thresholdType": 0,
"windowIntervalMs": 1000
},
"clusterMode": false,
"controlBehavior": 0,
"count": 1.0,
"durationInSec": 1,
"grade": 1,
"limitApp": "default",
"maxQueueingTimeMs": 0,
"paramFlowItemList": [],
"paramIdx": 1,
"resource": "hot"
}
}, {
"app": "mall-user-sentinel-rule-push-demo",
"gmtCreate": 1616137178470,
"gmtModified": 1616658923519,
"id": 2,
"ip": "192.168.3.1",
"port": 8719,
"rule": {
"burstCount": 0,
"clusterConfig": {
"fallbackToLocalWhenFail": true,
"sampleCount": 10,
"thresholdType": 0,
"windowIntervalMs": 1000
},
"clusterMode": false,
"controlBehavior": 0,
"count": 3.0,
"durationInSec": 1,
"grade": 1,
"limitApp": "default",
"maxQueueingTimeMs": 0,
"paramFlowItemList": [{
"classType": "int",
"count": 1,
"object": "4"
}],
"paramIdx": 0,
"resource": "findOrderByUserId"
}
}]
我提供两种解决思路:
1.自定义一个解析热点规则配置的解析器FlowParamJsonConverter,继承JsonConverter,重写convert方法。然后利用后置处理器替换beanName为"param-flow-rules-sentinel-nacos-datasource"的converter属性,注入FlowParamJsonConverter。
@Configuration
public class ConverterConfig {
@Bean("sentinel-json-param-flow-converter2")
@Primary
public JsonConverter jsonParamFlowConverter() {
return new FlowParamJsonConverter(new ObjectMapper(), ParamFlowRule.class);
}
}
@Component
public class FlowParamConverterBeanPostProcessor implements BeanPostProcessor {
@Autowired
private JsonConverter jsonParamFlowConverter;
@Override
public Object postProcessBeforeInitialization(Object bean, String beanName) throws BeansException {
if (beanName.equals("param-flow-rules-sentinel-nacos-datasource")) {
NacosDataSourceFactoryBean nacosDataSourceFactoryBean = (NacosDataSourceFactoryBean) bean;
nacosDataSourceFactoryBean.setConverter(jsonParamFlowConverter);
return bean;
}
return bean;
}
}
public class FlowParamJsonConverter extends JsonConverter {
Class ruleClass;
public FlowParamJsonConverter(ObjectMapper objectMapper, Class ruleClass) {
super(objectMapper, ruleClass);
this.ruleClass = ruleClass;
}
@Override
public Collection<Object> convert(String source) {
List<Object> list = new ArrayList<>();
JSONArray jsonArray = JSON.parseArray(source);
for (int i = 0; i < jsonArray.size(); i++) {
//解析rule属性
JSONObject jsonObject = (JSONObject) jsonArray.getJSONObject(i).get("rule");
Object object = JSON.toJavaObject(jsonObject, ruleClass);
list.add(object);
}
return list;
}
}
2. 改造Sentinel Dashboard控制台,发布配置时将ParamFlowRuleEntity转成ParamFlowRule类型,再发布到Nacos配置中心。从配置中心拉取配置后将ParamFlowRule转成ParamFlowRuleEntity。
从配置中心拉取配置到控制台时,FlowRule转换为FlowRuleEntity
从控制台发布配置到配置中心时,FlowRuleEntity转换为FlowRule
2. sentinel规则持久化部分源码分析
https://www.processon.com/view/link/607fef267d9c08283ddc2f8d