Spring 6中Spring Batch无数据库依赖实现方案咨询

方案一：自定义内存版Job Repository实现 #

既然官方移除了Map-based实现，你可以自己基于线程安全的内存集合实现一套符合Spring Batch要求的Job Repository组件，核心是实现JobRepository、JobExecutionDao、StepExecutionDao和ExecutionContextDao这几个接口，用ConcurrentHashMap存储作业执行数据，保证多线程场景下的安全性。

核心代码示例 #

先实现JobExecutionDao：

import org.springframework.batch.core.JobExecution;
import org.springframework.batch.core.repository.dao.JobExecutionDao;
import org.springframework.batch.core.repository.dao.NoSuchObjectException;

import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.atomic.AtomicLong;

public class InMemoryJobExecutionDao implements JobExecutionDao {
    private final ConcurrentMap<Long, JobExecution> jobExecutions = new ConcurrentHashMap<>();
    private final AtomicLong idGenerator = new AtomicLong(1);

    @Override
    public void saveJobExecution(JobExecution jobExecution) {
        if (jobExecution.getId() == null) {
            jobExecution.setId(idGenerator.getAndIncrement());
        }
        jobExecutions.put(jobExecution.getId(), jobExecution);
    }

    @Override
    public void updateJobExecution(JobExecution jobExecution) {
        if (!jobExecutions.containsKey(jobExecution.getId())) {
            throw new NoSuchObjectException("JobExecution not found: " + jobExecution.getId());
        }
        jobExecutions.put(jobExecution.getId(), jobExecution);
    }

    @Override
    public JobExecution getJobExecution(Long executionId) {
        JobExecution execution = jobExecutions.get(executionId);
        if (execution == null) {
            throw new NoSuchObjectException("JobExecution not found: " + executionId);
        }
        return execution;
    }

    // 其他接口方法按需实现（如findJobExecutions、getLastJobExecution等）
}

接着实现StepExecutionDao、JobInstanceDao、ExecutionContextDao的内存版本，逻辑类似，均用ConcurrentHashMap存储对应实体。最后配置自定义JobRepository：

import org.springframework.batch.core.repository.JobRepository;
import org.springframework.batch.core.repository.support.SimpleJobRepository;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.transaction.PlatformTransactionManager;

@Configuration
public class InMemoryBatchConfig {

    @Bean
    public JobRepository jobRepository(PlatformTransactionManager transactionManager) {
        return new SimpleJobRepository(
                new InMemoryJobExecutionDao(),
                new InMemoryStepExecutionDao(),
                new InMemoryJobInstanceDao(),
                new InMemoryExecutionContextDao()
        );
    }
}

方案二：禁用作业状态跟踪（限无状态批处理场景） #

如果你的批处理作业不需要跟踪执行状态（比如一次性运行、无需重试/恢复的简单任务），可以通过调整配置，让Spring Batch跳过状态持久化。不过这种方式仅适用于不需要作业历史、失败重试的场景：

配置示例 #

import org.springframework.batch.core.configuration.annotation.DefaultBatchConfigurer;
import org.springframework.batch.core.repository.JobRepository;
import org.springframework.context.annotation.Configuration;

@Configuration
public class NoTrackingBatchConfig extends DefaultBatchConfigurer {

    @Override
    protected JobRepository createJobRepository() throws Exception {
        return new JobRepository() {
            @Override
            public JobExecution createJobExecution(String jobName) {
                return new JobExecution(new JobInstance(1L, jobName));
            }

            // 其余接口方法按需空实现或返回默认值，比如save/update方法直接返回不做操作
        };
    }
}

注意事项 #

• 内存版实现不支持应用重启后的状态恢复，所有作业执行数据会在应用关闭后丢失；
• 仅支持单实例运行，不适合集群部署场景；
• 如果作业需要重试、暂停恢复或者需要保留执行历史，这种方案不适用，只能接受官方推荐的嵌入式数据库方案。

内容的提问来源于stack exchange，提问作者Rishabh