ElasticJob provides a customized Spring Boot Starter, which can be used in conjunction with Spring Boot. Developers are free from configuring CoordinatorRegistryCenter, JobBootstrap by using ElasticJob Spring Boot Starter. What developers need to solve distributed scheduling problem are job implementations with a little configuration.
The following content is only demonstrated through Spring Boot 3. The relevant content may still be valid on Spring Boot 2, but since Spring Boot 2 has ended maintenance, no availability assumptions are made for Spring Boot 2.
Job implementation is similar to other usage of ElasticJob. The difference is that jobs will be registered into the Spring IoC container.
Thread-Safety Issue
Bean is singleton by default.
Consider setting Bean Scope to prototype
if the instance of ElasticJob would be used by more than a JobBootstrap.
@Component
public class SpringBootDataflowJob implements DataflowJob<Foo> {
@Override
public List<Foo> fetchData(final ShardingContext shardingContext) {
// fetch data
}
@Override
public void processData(final ShardingContext shardingContext, final List<Foo> data) {
// process data
}
}
Configure the Zookeeper which will be used by ElasticJob via configuration files.
elasticjob.jobs
is a Map. Using key as job name. Specific job type and configuration in value.
The Starter will create instances of OneOffJobBootstrap
or ScheduleJobBootstrap
and register them into the Spring IoC container automatically.
Configuration reference:
elasticjob:
regCenter:
serverLists: localhost:6181
namespace: elasticjob-springboot
jobs:
dataflowJob:
elasticJobClass: org.apache.shardingsphere.elasticjob.dataflow.job.DataflowJob
cron: 0/5 * * * * ?
shardingTotalCount: 3
shardingItemParameters: 0=Beijing,1=Shanghai,2=Guangzhou
scriptJob:
elasticJobType: SCRIPT
cron: 0/10 * * * * ?
shardingTotalCount: 3
props:
script.command.line: "echo SCRIPT Job: "
Just start Spring Boot Starter directly. The schedule jobs will startup when the Spring Boot Application is started.
When to execute OneOffJob is up to you.
Developers can inject the OneOffJobBootstrap
bean into where they plan to invoke.
Trigger the job by invoking execute()
method manually.
Users should not use annotations such as jakarta.annotation.Resource
which partially violate Spring Boot best practices to inject Spring beans that define one-time tasks.
The bean name of OneOffJobBootstrap
is specified by property “jobBootstrapBeanName”,
Please refer to Spring Boot Starter Configuration.
elasticjob:
jobs:
myOneOffJob:
elasticJobType: SCRIPT
jobBootstrapBeanName: myOneOffJobBean
shardingTotalCount: 9
props:
script.command.line: "echo Manual SCRIPT Job: "
import org.apache.shardingsphere.elasticjob.bootstrap.type.OneOffJobBootstrap;
import org.springframework.beans.factory.ObjectProvider;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;
import java.util.Objects;
@RestController
public class OneOffJobController {
// Injection via "@Autowired"
@Autowired
@Qualifier("myOneOffJobBean")
private ObjectProvider<OneOffJobBootstrap> myOneOffJobProvider;
@GetMapping("/execute2")
public String executeOneOffJob2() {
OneOffJobBootstrap myOneOffJob = myOneOffJobProvider.getIfAvailable();
Objects.requireNonNull(myOneOffJob);
myOneOffJob.execute();
return "{\"msg\":\"OK\"}";
}
}
In the process of using ElasticJob, when the job is abnormal, the following error handling strategies can be used.
Error handler strategy name | Description | Built-in | Default | Extra config |
---|---|---|---|---|
Log Strategy | Log error and do not interrupt job | Yes | Yes | |
Throw Strategy | Throw system exception and interrupt job | Yes | ||
Ignore Strategy | Ignore exception and do not interrupt job | Yes | ||
Email Notification Strategy | Send email message notification and do not interrupt job | Yes | ||
Wechat Enterprise Notification Strategy | Send wechat message notification and do not interrupt job | Yes | ||
Dingtalk Notification Strategy | Send dingtalk message notification and do not interrupt job | Yes |
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: LOG
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: THROW
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: IGNORE
Please refer to here for more details.
Maven POM:
<dependency>
<groupId>org.apache.shardingsphere.elasticjob</groupId>
<artifactId>elasticjob-error-handler-email</artifactId>
<version>${latest.release.version}</version>
</dependency>
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: EMAIL
props:
email:
host: host
port: 465
username: username
password: password
useSsl: true
subject: ElasticJob error message
from: from@xxx.xx
to: to1@xxx.xx,to2@xxx.xx
cc: cc@xxx.xx
bcc: bcc@xxx.xx
debug: false
Please refer to here for more details.
Maven POM:
<dependency>
<groupId>org.apache.shardingsphere.elasticjob</groupId>
<artifactId>elasticjob-error-handler-wechat</artifactId>
<version>${latest.release.version}</version>
</dependency>
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: WECHAT
props:
wechat:
webhook: you_webhook
connectTimeout: 3000
readTimeout: 5000
Please refer to here for more details.
Maven POM:
<dependency>
<groupId>org.apache.shardingsphere.elasticjob</groupId>
<artifactId>elasticjob-error-handler-dingtalk</artifactId>
<version>${latest.release.version}</version>
</dependency>
elasticjob:
regCenter:
...
jobs:
...
jobErrorHandlerType: DINGTALK
props:
dingtalk:
webhook: you_webhook
keyword: you_keyword
secret: you_secret
connectTimeout: 3000
readTimeout: 5000