12c SOA Weblogic - how to apply patch in weblogic server

Steps for applying patch:

Step1: Stop all servers and clean out the domain_home/servers/<server>/ 'tmp', 'cache' and 'stage' folders.

Step2: Set oracle home and test the opatch version. Set the ORACLE_HOME environment variable to the directory where you have installed Oracle SOA Suite.
cd /app/Oracle/Middleware/Oracle_Home/OPatch  export ORACLE_HOME=/app/Oracle/Middleware/Oracle_Home

opatch version 

opatch lsinventory

Step3: Copy patch file and unzip the .zip folder in unix directory like /soashare/patch.
Copy downloaded patch file in both node of server /tmp/soadev directory using WinScp 

• Change the permission of downloaded patch

                chmod 777 p22337707_122100_Generic.zip 

•       Move downloaded patch from /tmp/soadev directory to /soashare/patch directory

                mv p22337707_122100_Generic.zip /soashare/patch
                cd /soashare/patch

          chmod 777 p22337707_122100_Generic.zip

•  Unzip the patch zip file into the PATCH_TOP folder.

Note: Rename existing “PATCH_TOP” folder in /soashare/patch to “PATCH_TOP_<<Name of last patch applied>>”

unzip -d PATCH_TOP p22337707_122100_Generic.zip

Step4: Apply patch  

    Set your current directory to the directory where the patch is located.

    cd PATCH_TOP/22337707       

    $ORACLE_HOME/OPatch/opatch apply

Step5: Check the applied patch

   $ORACLE_HOME/OPatch/opatch lsinventory

Step6: Follow above steps 2-4 in node 2.

Step7: Start all Servers.

++END++

Patch applied in production:

1. Patch 22337707: Naming conflict with EDN event subscribers from single Mediator component

SYMPTOMS: When you are deploying a composite defined with one Mediator and 5 events subscriptions you are expecting durable subscriptions (JMS) to 5 events. Instead only 3 subscriptions are created so 2 events aren't "read".

CAUSE: Bug 22337707 - Deploying composite doesn't create enough durable subscribers (JMS) for EDN.

The problem seems to be related to a naming convention problem. The event Subscription Name in JMS is comprised of the first 10 characters of the Event Name + the Instance ID.

For example:

Abcdefghij_SCA_220103

So in the case of the two events below, only one subscription can be created due to a naming conflict:

AbcdefghijklEventOne

AbcdefghijklEventTwo

Reference doc:

Composite deployment fails to create enough durable subscribers (JMS) for EDN (Doc ID 2091244.1)

2. Patch 22385445: Flow Trace working too long or not loading at all

SYMPTOMS: When selecting the Flow ID in the EM console, the audit trail page hangs and it either takes a while to appear or it never appears at all.
CAUSE: Issue due to Bug: 22385445 12.2.1.0.0 Flow Trace working too long or not loading at all
Reference doc:
SOA 12c: It takes a very long time to display the flow trace in EM or it fails to display at all (Doc ID 2208520.1)

3. Patch 23138916: MERGE REQUEST ON TOP OF 12.2.1.0.0 FOR BUGS 21549249 21572567
Bugs Resolved by This Patch
21549249      cannot insert NULL into SOAINFRA.AUDIT_DETAILS.CI_PARTITION_DATE

21572567      composite instance id not restored during non-SOA ws callback

4. Patch 25388847: WLS PATCH SET UPDATE 12.2.1.0.170418 (Supersede of patch 24904865)
Bugs Resolved by This Patch
21495475      Fix for Bug 21495475
22049932      Fix for Bug 22049932
22175246      Fix for Bug 22175246
22200449      Fix for Bug 22200449
22200491      Fix for Bug 22200491
22200523      Fix for Bug 22200523
22200594      Fix for Bug 22200594
22241704      TRACKING BUG FOR BACKPORT BUG 21621232
22247869      Fix for Bug 22247869
22248079      Fix for Bug 22248079
Reference Document:

SOA 12.2.1 Enterprise Manager Latency (Doc ID 2240093.1)

5. Patch 23762529: MT: STRESS:WLS: MBEANS TAKING MORE TIME
Bugs Resolved by This Patch

23762529      MT: STRESS:WLS: MBEANS TAKING MORE TIME

6. Patch 24327938: TOPLINK SECURITY PATCH UPDATE CPUJUL2016
Bugs Resolved by This Patch

23630882        Fix for Bug 23630882

7. Patch 22090514: "new code mTestMeadataMgr=null" data in server out file
Reference Document:

new code mTestMeadataMgr=null", in the SOA server .out log (Doc ID 2104408.1)

8. SOA 12.1.3: Undo tablespace becomes exhausted during purge operations (Doc ID 2224031.1)
Bug 24558237 - Purge causing'ORA-01652 or 'Undo tablespace full' on DOCUMENT_DLV_MSG_REF

determined this was a problem on the scripts