JDBC Inbound Adapter fails on update with exception
I am using Spring version 3.0.5 with Integration version 2.0.3 (I have also tried with 2.0.5 and received the same exception). I am doing the local development work on a Windows 7 machine with Java 1.6.0_11 installed. I am seeing the exception on a testing host that I use that is running RHEL 5.6 with Java 1.6.0_16 installed.
The software is basically pulling some records from a database, doing some basic processing on them, and pushing them out to a JMS destination.
My configuration is something like this:
Code:
<int-jdbc:inbound-channel-adapter id="readDatabaseResults"
channel="databaseResults"
data-source="datasource"
row-mapper="dataResultRowMapper"
max-rows-per-poll="50"
query="select record_id, record_date, record_value from my_table where completed is null"
update="update my_table set completed = sysdate where record_id = :recordId and record_date = :recordDate"
update-per-row="true">
<poller fixed-rate="5000">
<transactional/>
</poller>
</int-jdbc:inbound-channel-adapter>
<channel id="databaseResults"/>
<splitter id="splitResults"
input-channel="databaseResults"
output-channel="individualDatabaseResults"
expression="payload"/>
<channel id="individualDatabaseResults"/>
<service-activator id="databaseResultProcessor"
input-channel="individualDatabaseResults"
output-channel="processedDatabaseResults"
ref="springResultProcessor"
method="processResultRecords"/>
<channel id="processedResults"/>
<transformer id="resultToMessageTransformer"
ref="resultTransformer"
method="transformResult"
input-channel="processedResults"
output-channel="processedResultMessages"/>
<channel id="processedResultMessages"/>
<int-jms:outbound-channel-adapter id="resultMessageQueue"
channel="processedResultMessages"
connection-factory="jmsConnectionFactory"
destination="resultDestination"
message-converter="myMessageConverter"/>
After some time (approximately 900-1000 records, though this number seems to vary making it even harder to track down this problem it seems), I receive the following exception, which appears to be coming from the parameter mapping during the update attempt after reading and mapping the results to my data type. Once this failure occurs, it will re-occur with every poll going forward. It is a pair of exceptions so I will seperate the log lines for readability (see next post)
This problem only seems to occur on the RHEL host as I have not been able to reproduce it yet on my Windows machine. I have also tried verifying that it is not an issue with the record by having it process the record it failed on alone. It also seems to have no issue processing the record it failed on later after I restart the process.
