SYMPTOM:
After successfully installing OIG(RCU,SOA,IDM,offlineconfig), weblogic server started. When we tried to start SOA , we have seen error logs related to DB as attached. Somehow, SOA seemed to be in state of RUNNING eventhough there were DB related erros in logs. When we tried to start OIM afterwards, same errors showed up in logs and OIM server failed to start.
OIG logs:
------------------------------------------------------
INFO: Jps is reset.
Nov 22, 2018 10:33:02 AM oracle.security.jps.internal.policystore.util.JpsDataManagerUtil getDBConnection
WARNING: Can not connect DB with URL jdbc:oracle:thin:@//x.x.x.x:1521/sid_name
java.sql.SQLRecoverableException: IO Error: Got minus one from a read call, connect lapse 32 ms., Authentication lapse 0 ms.
at oracle.jdbc.driver.T4CConnection.logon(T4CConnection.java:795)
Caused by: java.io.IOException: Got minus one from a read call, connect lapse 32 ms., Authentication lapse 0 ms.
at oracle.jdbc.driver.T4CConnection.logon(T4CConnection.java:791)
Caused by: java.io.IOException: Got minus one from a read call, connect lapse 32 ms.
at oracle.net.ns.NSProtocolNIO.negotiateConnection(NSProtocolNIO.java:147)
Caused by: oracle.net.ns.NetException: Got minus one from a read call
at oracle.net.ns.NIOHeader.readHeaderBuffer(NIOHeader.java:101)
Nov 22, 2018 10:33:02 AM oracle.security.jps.internal.policystore.util.JpsDataManagerUtil getDBConnection
WARNING: Can not connect DB with URL jdbc:oracle:thin:@//10.11.96.74:1521/TKACSMNG
java.sql.SQLRecoverableException: IO Error: Got minus one from a read call, connect lapse 23 ms., Authentication lapse 0 ms.
Nov 22, 2018 10:33:02 AM oracle.security.jps.internal.credstore.ldap.LdapCredentialStore init
WARNING: Could not create credential store instance. Reason oracle.security.jps.service.policystore.PolicyStoreConnectivityException: oracle.security.jps.service.policystore.PolicyStoreException: Can not connect DB with URL [jdbc:oracle:thin:@//-------:1521/-------]
Nov 22, 2018 10:33:02 AM oracle.security.jps.internal.config.OpssCommonStartup start
INFO: Jps startup failed.
<Nov 22, 2018 10:33:02,733 AM EET> <Critical> <WebLogicServer> <BEA-000362> <Server failed. Reason:
<Nov 22, 2018 10:33:02,795 AM EET> <Notice> <WebLogicServer> <BEA-000365> <Server state changed to FAILED.>
<Nov 22, 2018 10:33:02,795 AM EET> <Error> <WebLogicServer> <BEA-000383> <A critical service failed. The server will shut itself down.>
<Nov 22, 2018 10:33:02,809 AM EET> <Notice> <WebLogicServer> <BEA-000365> <Server state changed to FORCE_SHUTTING_DOWN.>
Stopping Derby server...
Oracle DB logs:
------------------------------------------------------
ORA-00020: maximum number of processes (500) exceeded
ORA-20 errors will not be written to the alert log for
CAUSE:
DB side, maximum number of processes max limit has been exceeded.
Does not exists on OIM 11G R2. Support confirmed the situation. As per their description; Because the FMW 12c use database as repository for all components, the number of connections and processes versus 11g platform was increased. In the old platform the recommendation was very similar for starting: processes 500 and open_cursors 800 ! But in 12c the supplementary connections for WLS, MDS and OPSS contribute to the general increased necessary resources. Looks like is a doc bug in documentation, because the same value was maintained in the documentation of 12c.
RESOLUTION:
The max.process limit was 500. They increased it to 1000 and the issue got resolved. Recommended values:
processes 1000
open_cursors 1600
The situation is confirmed. FOLLOW-UP ON:
Oracle Support Bug 29138957 (REPOSITORY CREATION UTILITY REQUIREMENTS INITIAL VALUE) can be found at: https://support.oracle.com/epmos/faces/BugDisplay?id=29138957