Time Out Settings in OSB

Question

Hi All,
I have a service which calls DB via DB adapter. For the webservice call, I have an option of read time out settings in the business service, so that it will get time out after the configured time period.
But for the DB call I dont have this option in the business service and we are getting stuck thread exceptions when the legacy is taking more time to respond back. We have a requirement that all the services realted to DB call have to get time out when the legacy takes more time.
Kindly help me in doing this.
Thanks and Regards,
Prabhu

Answer 1

Hi Anuj,
I did a sample stored procedure which will wait for number of seconds which we send as input and will send the success repose after the number of seconds.
StoredProcedure:
create or replace
procedure sleep (seconds in number,response out varchar2) as testdata number;
begin
testdata := seconds;
dbms_lock.sleep(testdata);
response := 'SUCCESS';
end;
Business Service XML:
<?xml version="1.0" encoding="UTF-8"?>
<xml-fragment xmlns:ser="http://www.bea.com/wli/sb/services" xmlns:tran="http://www.bea.com/wli/sb/transports" xmlns:env="http://www.bea.com/wli/config/env" xmlns:jca="http://www.bea.com/wli/sb/transports/jca" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<ser:coreEntry isProxy="false" isEnabled="true">
<ser:binding type="SOAP" isSoap12="false" xsi:type="con:SoapBindingType" xmlns:con="http://www.bea.com/wli/sb/services/bindings/config">
<con:wsdl ref="Sleep2/Sleep_dbBS"/>
<con:binding>
<con:name>Sleep_ptt-binding</con:name>
<con:namespace>http://xmlns.oracle.com/pcbpel/adapter/db/Application1/Project2/Sleep</con:namespace>
</con:binding>
</ser:binding>
<ser:monitoring isEnabled="false">
<ser:aggregationInterval>10</ser:aggregationInterval>
<ser:pipelineMonitoringLevel>Pipeline</ser:pipelineMonitoringLevel>
</ser:monitoring>
<ser:reporting>true</ser:reporting>
<ser:sla-alerting isEnabled="true">
<ser:alertLevel>normal</ser:alertLevel>
</ser:sla-alerting>
<ser:ws-policy>
<ser:binding-mode>wsdl-policy-attachments</ser:binding-mode>
</ser:ws-policy>
</ser:coreEntry>
<ser:endpointConfig>
<tran:provider-id>jca</tran:provider-id>
<tran:inbound>false</tran:inbound>
<tran:URI>
<env:value>jca://eis/DB/SleepConnection1</env:value>
</tran:URI>
<tran:outbound-properties>
<tran:load-balancing-algorithm>round-robin</tran:load-balancing-algorithm>
<tran:retry-count>0</tran:retry-count>
<tran:retry-interval>30</tran:retry-interval>
<tran:retry-application-errors>true</tran:retry-application-errors>
</tran:outbound-properties>
<tran:all-headers>false</tran:all-headers>
<tran:provider-specific>
<jca:jca-file ref="Sleep2/Sleep_db"/>
<jca:adapter-name>Sleep</jca:adapter-name>
<jca:adapter-type>DATABASE</jca:adapter-type>
<jca:always-use-wsdl>true</jca:always-use-wsdl>
<jca:connection-mode>managed</jca:connection-mode>
<jca:outbound-properties>
<jca:operation-properties>
<jca:operation-name>Sleep</jca:operation-name>
<jca:spec-properties>
<jca:property>
<jca:name>ProcedureName</jca:name>
<jca:value>SLEEP</jca:value>
</jca:property>
<jca:property>
<jca:name>QueryTimeout</jca:name>
<jca:value>2</jca:value>
</jca:property>
<jca:property>
<jca:name>GetActiveUnitOfWork</jca:name>
<jca:value>false</jca:value>
</jca:property>
</jca:spec-properties>
</jca:operation-properties>
</jca:outbound-properties>
</tran:provider-specific>
</ser:endpointConfig>
</xml-fragment>
When I run this from the business service, if I give the input as 10, we are getting error only after 10 seconds not in the query time out seconds. But the thing is we are getting error. If we set the time as 11 seconds we are getting success response. But our requirement is the procedure has to time out in the specified seconds. Kindly correct me if I am wrong.
Regards,
Prabhu