Account lock error when opmnctl start

Question

Hi
After a longer time I tried to reboot my Portal 10.
When running opmnctl startall, the service was hanging and only starting the following:
------------------------------------------------+---------
ias-component | process-type | pid | status
------------------------------------------------+---------
LogLoader | logloaderd | N/A | Down
dcm-daemon | dcm-daemon | N/A | Down
DSA | DSA | N/A | Down
OC4J | OC4J_SECURITY | N/A | Down
HTTP_Server | HTTP_Server | 12205 | Alive
OID | OID | 12210 | Alive
The logfile on /app/oracle/product/infra10g/opmn/logs/SIDA shows me following:
07/06/18 18:14:45 Dependency check
Error - ORA-28000: the account is locked
SIDA is available
SUCCEEDED
Now, I can not find any locked accounts.
I tried to run oidadmin, but it tells me the server is not up.
Do you have any clue where I could unlock the account?
Do you know where I can check which account is locked?
Thanks in advance!
Reto

Answer 1

After getting in contact with Oracle SR the problem could get resolved.
Here the log and solution:
running programm:
on infra environment: opmnctl startall
Issue:     
Error - ORA-28000: the account is locked
Can not start OC4J_SECURITY
Logfile found on: /app/oracle/product/infra10g/opmn/logs/TZHSRRE1
Solution:
We shouldn't worry about the ORA-28000, ORA-28001 errors - these are misleading errors caused by the "scott" DB account that it's not used by OID/SSO.
The real problem appears to be the OC4J_SECURITY process being forcefully stopped/crashing.
First of all, let's check id the OID is working as expected:
Check the ports on /app/oracle/product/comp/infra10g/install/portlist.ini
1. ps -ef | grep oidldapd
2. $ORACLE_HOME/bin/ldapbind -h <oid_host> -p <oid_nonSSL_port>
-> $ORACLE_HOME/bin/ldapbind -h zus95d-4403.tszrh.csfb.com -p 389
Cannot connect to the LDAP server
3. $ORACLE_HOME/bin/ldapbind -h <oid_host> -p <oid_SSL_port> -U 1
-> $ORACLE_HOME/bin/ldapbind -h zus95d-4403.tszrh.csfb.com -p 636 -U 1
Cannot connect to the LDAP server
Stop and restart OID on this way:
4. $ORACLE_HOME/opmn/bin/opmnctl stopproc process-type=OID
5. check if there are still running oidldap processes ps -ef | grep oidldapd
In this case just kill them
6. Connect as SYS user to the Metadata repository database and issue the following statements:
+ delete ods.ods_process where serverid=2;
+ commit;
7. $ORACLE_HOME/opmn/bin/opmnctl startproc process-type=OID
8. after some minutes check OID again:
$ORACLE_HOME/bin/ldapbind -h zus95d-4403.tszrh.csfb.com -p 389
bind successful
$ORACLE_HOME/bin/ldapbind -h zus95d-4403.tszrh.csfb.com -p 636 -U 1
bind successful
9. Try to start OC4J_SECURITY:
$ORACLE_HOME/opmn/bin/opmnctl startproc process-type=OC4J_SECURITY
10. Try to login on SSO:
http://servername.host:port/portal/page/portal/TOPLEVELSITE/Welcome
11. Connection works
Note/Conclustion:
- The Oracle processes should be always cleanly stopped before rebooting the server
- You should access the OID tables (from ODS schema) only if requested by the Oracle Support.
- serverid=2 is the entry corresponding to the OID Server process (oidldapd)
- serverid=7 is the entry corresponding to the OID Sinchronization and Provisioning Server.
- The "state=2" means "up and running", "state=1" means "starting", "state=4" means "stopped".
ods_process table:
INSTANCE     PID          SERVERID     FLAGS               HOSTNAME     STATE     LASTWAKETIME          RETRYCOUNT
1.00          3'997.00     7.00          instance=1 configset=0      "SERVERNAME"     2.00     06.18.2007 18:13:16     0.00
1.00          3'996.00     2.00          -i 1 -conf 0          "SERVERNAME"     2.00     06.19.2007 17:28:47     0.00