By Clemens Bleile

Recently I upgraded an ODA X7-M from 19.12. to 19.13. After the dcs-, server- and storage-upgrade several databases on the machine had to be patched from 19.9. to 19.13. The

[[email protected]<node> ~]# odacli create-prepatchreport --dbhome --dbhomeid <home-id> -v 19.13.0.0.0

went through without reporting an issue, but during the

[[email protected]<node> ~]# odacli update-dhome -i <home-id> -v 19.13.0.0.0 -f

I got an error “DCS-10001:Internal error encountered: null.”:

[[email protected]<node> ~]# odacli describe-job -i "<Job-Id>"

Job details
----------------------------------------------------------------
                     ID:  <Job-Id>
            Description:  DB Home Patching: Home Id is <Home-Id>
                 Status:  Failure
                Created:  February 13, 2022 1:28:41 PM CET
                Message:  DCS-10001:Internal error encountered: null.

Task Name                                Start Time                          End Time                            Status
---------------------------------------- ----------------------------------- ----------------------------------- ----------
DB Home Patching                         February 13, 2022 1:29:02 PM CET    February 13, 2022 1:33:28 PM CET    Failure
DB Home Patching                         February 13, 2022 1:29:02 PM CET    February 13, 2022 1:33:28 PM CET    Failure
Adding USER SSH_EQUIVALENCE              February 13, 2022 1:29:02 PM CET    February 13, 2022 1:29:02 PM CET    Success
Adding USER SSH_EQUIVALENCE              February 13, 2022 1:29:02 PM CET    February 13, 2022 1:29:03 PM CET    Success
Adding USER SSH_EQUIVALENCE              February 13, 2022 1:29:03 PM CET    February 13, 2022 1:29:04 PM CET    Success
task:TaskSequential_3705                 February 13, 2022 1:29:04 PM CET    February 13, 2022 1:33:04 PM CET    Failure
Creating wallet for DB Client            February 13, 2022 1:29:44 PM CET    February 13, 2022 1:29:44 PM CET    Success
Patch databases by RHP                   February 13, 2022 1:29:44 PM CET    February 13, 2022 1:31:28 PM CET    Success
updating database metadata               February 13, 2022 1:32:22 PM CET    February 13, 2022 1:32:22 PM CET    Success
Set log_archive_dest for Database        February 13, 2022 1:32:22 PM CET    February 13, 2022 1:32:25 PM CET    Success
Patch databases by RHP                   February 13, 2022 1:32:25 PM CET    February 13, 2022 1:33:04 PM CET    Failure

[[email protected]<node> ~]#

Checking the RHP logfile /opt/oracle/rhp/rhplog/rhpapi.log.0 showed the following:

[UID:<UID>] [Patch databases by RHP : JobId=<Job-Id>] [ 2022-02-13 13:33:04.671 CET ] [BatchMoveOpImpl.internalContinueMove:3348]  Batch failed for at least one DB
[UID:<UID>] [Patch databases by RHP : JobId=lt.Job-Id>] [ 2022-02-13 13:33:04.671 CET ] [DBPatchUpgradeOperationImpl.move:477]  attempt to move or upgrade database failed with OperationException : PRCT-1003 : failed to run "rhphelper" on node "<node>"
PRCT-1014 : Internal error: RHPHELP112_mergeLsnr-08
[UID:<UID>] [Patch databases by RHP : JobId=lt.Job-Id>] [ 2022-02-13 13:33:04.671 CET ] [DatabaseOperationImpl.move:1448]  OperationException: PRCT-1003 : failed to run "rhphelper" on node "<node>"
PRCT-1014 : Internal error: RHPHELP112_mergeLsnr-08
[UID:<UID>] [Patch databases by RHP : JobId=lt.Job-Id>] [ 2022-02-13 13:33:04.671 CET ] [GHOperationCommonImpl.moveDatabase:2978]  OperationException: PRCT-1003 : failed to run "rhphelper" on node "<node>"
PRCT-1014 : Internal error: RHPHELP112_mergeLsnr-08
[UID:<UID>] [Patch databases by RHP : JobId=lt.Job-Id>] [ 2022-02-13 13:33:04.671 CET ] [FPPMBeanImpl.doOp:372]  InvocationTargetException caught
[UID:<UID>] [Patch databases by RHP : JobId=lt.Job-Id>] [ 2022-02-13 13:33:04.672 CET ] [FPPMBeanImpl.doOp:382]  Exception:
java.lang.NullPointerException: null

So basically the errors were

PRCT-1003 : failed to run "rhphelper" on node "<node>"
PRCT-1014 : Internal error: RHPHELP112_mergeLsnr-08

The analysis revealed the root cause for the issue to be Bug 32833813. This is described in My Oracle Support Note 32833813.8 for an Upgrade from 11.2. to 19.2., but it obviously may also happen when moving to a new ORACLE_HOME for an Release Update with RHP:

Upgrading 11.2 SI Database to 19c Failed With Error PRCT-1003 : failed to run “rhphelper” on node “<HOSTNAME>” PRCT-1014 : internal error: rhphelp12102_main-02

The bug is fixed in the 19.14.0.0.220118 (JAN 2022) OCW Release Update. According the MOS Note there is no workaround.

The workaround for me was to change the ORACLE_HOME in the cluster registry to 19.13. for all DBs and finally update the ODA registry. I.e. in my case with the 19.13.-ORACLE_HOME being /u01/app/odaorahome/oracle/product/19.0.0.0/dbhome_3:

[[email protected]<node> ~]$ srvctl stop database -db <DBNAME>
[[email protected]<node> ~]$ srvctl modify database -db <DBNAME> -oraclehome /u01/app/odaorahome/oracle/product/19.0.0.0/dbhome_3
[[email protected]<node> ~]$ srvctl config database -db <DBNAME>
[[email protected]<node> ~]$ srvctl start database -db <DBNAME>
[[email protected]<node> ~]$ . oraenv
ORACLE_SID = [oracle] ? <ORACLE_SID>
[[email protected]<node> ~]$ cd $ORACLE_HOME/OPatch
[[email protected]<node> ~]$ ./datapatch -verbose

REMARKs:
– After modifying the ORACLE_HOME with srvctl the agent automatically updates the /etc/oratab.
– On a standby-DB it’s of course not necessary to run datapatch.

To automate this I wrote a bash-script, so that I can start this against all my databases on this server running in the 19.9.-ORACLE_HOME.

The last step is to update the ODA registry to reflect the change of the databases ORACLE_HOME:

[[email protected]<node> ~]# odacli update-registry -n db -f

Summary: Sometimes it’s necessary to manually patch databases to a newer RU on the ODA without using odacli. Fortunately we have the command “odacli update-registry” to update the ODA registry concerning our manual change.


Thumbnail [60x60]
by
Clemens Bleile