Which threadsafe JobRepository for multithreaded Step? - multithreading

Question

While running spring batch job with partition, each of the slave threads are failing with the below error :
[SimpleAsyncTaskExecutor-2] AbstractStep.execute(250) | Encountered an error saving batch meta data. This job is now in an unknown state and should not be restarted.
org.springframework.dao.OptimisticLockingFailureException: Attempt to update step execution id=4623 with wrong version (1), where current version is 2 at org.springframework.batch.core.repository.dao.JdbcStepExecutionDao.updateStepExecution(JdbcStepExecutionDao.java:185) at org.springframework.batch.core.repository.support.SimpleJobRepository.update(SimpleJobRepository.java:171)
Job Repository being used is JobRepositoryFactoryBean. Spring batch version is 2.1.8 .
In BATCH_STEP_EXECUTION table, for all steps the version is 2, even while running the steps for the first time.
Job Configuration as below:
<job id="partitionJob" xmlns="http://www.springframework.org/schema/batch" parent="simpleJob">
<step id="masterStep" xmlns="http://www.springframework.org/schema/batch" parent="skipLimitStep">
<partition step="slave" partitioner="BulkPartitioner" xmlns="http://www.springframework.org/schema/batch">
<handler grid-size="10" task-executor="asyncTaskExecutor" xmlns="http://www.springframework.org/schema/batch" />
</partition>
</step>
</job>
<step id="slave" xmlns="http://www.springframework.org/schema/batch" parent="skipLimitStep">
<tasklet>
<chunk reader="partitionPagingItemReader" writer="partitionItemWriter" commit-interval="1" processor="partitionitemProcessor" skip-limit="1000">
</chunk>
<transaction-attributes propagation="REQUIRES_NEW" />
</tasklet>
</step>
<bean id="BulkPartitioner" class="foo.bar.BulkPartitioner" scope="step" >
<property name="model" value="#{jobParameters['abc']}"/>
</bean>
<bean id="partitionPagingItemReader" class="org.springframework.batch.item.database.JdbcPagingItemReader" scope="step" >
<property name="dataSource" ref="dataSource" />
<!-- <property name="fetchSize" value="#{jobParameters['fetch.size']}"></property> -->
<property name="fetchSize" value="1000"></property>
<property name="queryProvider">
<bean class="org.springframework.batch.item.database.support.SqlPagingQueryProviderFactoryBean">
<property name="dataSource" ref="dataSource" />
<property name="selectClause" value="select a, b, c" />
<property name="fromClause" value="from d" />
<property name="whereClause" value="where e= :f" />
<property name="sortKey" value="e" />
</bean>
</property>
<!-- Inject via the ExecutionContext in rangePartitioner -->
<property name="parameterValues">
<map>
<entry key="f" value="#{stepExecutionContext[f]}" />
</map>
</property>
<property name="pageSize" value="1" />
<property name="rowMapper">
<bean class="foo.bar.mapper" />
</property>
</bean>
<bean id="simpleJob"
class="org.springframework.batch.core.job.SimpleJob" abstract="true">
<property name="jobRepository" ref="jobRepository"/>
<property name="restartable" value="true"/>
</bean>
<bean id="jobRepository"
class="org.springframework.batch.core.repository.support.JobRepositoryFactoryBean">
<property name="dataSource" ref="dataSource"/>
<property name="databaseType" value="oracle"/>
<property name="transactionManager" ref="transactionManager"/>
<property name="isolationLevelForCreate" value="ISOLATION_DEFAULT"/>
<property name="validateTransactionState" value="false"/>
</bean>
update: This is happening only for failed steps. successful completion of steps are updated correctly.
What could be the cause? TIA.

Answer 1

In case this is a problem for others, this was the solution I found:
<property name="saveState" value="false" />
in the reader(s)/writer(s) for the concurrent job/step. As the logging mentions, "This job is now in an unknown state and should not be restarted" so you should consider:
<property name="restartable" value="false"/>
Sources:
http://docs.spring.io/spring-batch/reference/html/scalability.html
http://docs.spring.io/spring-batch/reference/html/readersAndWriters.html#process-indicator