BAWUpgradeManual
BAWUpgradeManual
SBP001
VERSION: 1.0
29-Dec-2020
Prepared by:
Revision Chart
The HADR replication of BPM Production database to DR database should be stopped before
taking backup. This step should be done on Friday 7 PM.
Backups:
This step should be done on Friday 8 PM.
Setups:
This step should be done on Wednesday 11 PM.
Following setup files should be present on BPM Production and DR server at designated directory:
BAW_20_0_0_1_Linux_x86_3_of_3.tar.gz
BAWE_20_0_0_1_Linux_x86_1_of_3.tar.gz
BAWE_20_0_0_1_Linux_x86_2_of_3.tar.gz
workflow.20001.delta.repository.zip
8.6.20020001-WS-BPM-IFJR62453.zip
8.6.20020001-WS-BPM-IFJR62480.zip
8.6.20020001-WS-BPM-IFJR62974.zip (this file needs to be re-downloaded from following link:
https://www.ibm.com/support/fixcentral/swg/downloadFixes?parent=ibm
%2FWebSphere&product=ibm/WebSphere/
IBM+Business+Automation+Workflow&release=All&platform=All&function=fixId&fixids=8.6.200
20001-WS-BPM-
IFJR62974&includeRequisites=1&includeSupersedes=0&downloadMethod=http)
Miscellaneous:
It is mandatory to have GUI Access enabled on BPM Production and DR Application servers. This
should be verified by Wednesday 4 PM.
The OAW authenticator application should be stopped from ECM Production WebSphere. This
should be verified by Friday 6 PM.
Since the BPM Process portal is still accessible to users from direct hostname URL, the users can
still access the application even if OAW application is stopped. It is recommended to block the
end user access by restricting following access from the user machines:
o Pr-bpmapp.sbp.org.pk 9081, 9444
o Dr-bpmapp.sbp.org.pk 9081, 9444
It is important to note that this access should not be blocked on DBA user machines, as they
would need to access these machines during upgrade. Further, it is also important to note that
this Port block is temporary and should be enabled once the upgrade activity is completed. This
should be verified by Friday 6 PM.
Below are the different fail-over scenarios and their respective resolutions. It is important to note that
these considerations are built on consideration that BPM application would be shut down for business
users at 6:PM Friday evening, and then we have the backups available as described in the Backup
section. It is also important to note that below scenarios also consider that ITSPMD has taken
application downtime from Friday 6:00 PM to Monday 6:00 AM.
If BAW upgrade fails during the upgrade activity, first we would try to resolve the problem by
analyzing the issue and tracing out log files. We would also raise PMR with Severity 1 to IBM
Support to get their urgent support. The cut of time to stop looking into resolution is
recommended to be 3:PM Sunday Afternoon. After that, we should proceed with restoring the
backups. Before restoring, the BPM log files should be saved for analyzing the problem later.
If BAW upgrade fails once business has resumed to normal on Monday morning, first we would
try to resolve the problem by analyzing the issue and tracing out log files. We would also raise
PMR with Severity 1 to IBM Support to get their urgent support. It is important to note that
since business cases would be down, we would need a cut-off time given by ITSPMD to stop
looking into resolution and move for restoration. We recommend to have 2 hours as cut-off time
for the efforts to resolve the problem. In case we go ahead with backup restore, it is important
to note that the business data from 9:00 AM in the morning would be lost, and once the backup
is restored, this data would need to resubmitted on the application.
DR Environment Upgrade:
In Production, we have HADR environment setup, where a separate BPM environment is present on DR
server, and its database is present as Standby mode. Both Production and DR environment are currently
on 8.6.201803, and we plan to upgrade both of them to 20.0.0.1.
The upgrade steps involve database scripts execution. During UAT upgrade, we observed that the scripts
upgrade existing databases (BPMDB, CMNDB etc.), and also create new database CPEDB. We
recommend following approach:
1. Stop HADR from Primary to Secondary environment. Turn on the BPM Services on DR site, and
verify that Process Portal and Admin Console are working correctly. This is to verify that DR site
is in working state before upgrade.
2. Perform all the upgrade steps in parallel on Production and DR site. All the steps including
database scripts step should be executed on both Production and DR.
3. Once the upgrade is successful on Production and DR, we need to verify both environments
after starting services. Both environments should be in working state.
4. Now, we need to setup HADR of all databases (BPMDB, CMNDB, PDWDB, KM_ECM, CPEDB)
from scratch (Restore cold backup from Production on DR database and then setup HADR).
5. Once HADR is successfully setup, we need to verify the application Production and then switch
over to DR environment to verify the application from DR site.
/bpmapp/bpmuser/UpgradeMedia
Cd /bpmapp/bpmuser/UpgradeMedia/IM64
./userinst
Click on restart IM to verify the verion.
- After successful update, run the installation manager from below command:
cd IBM/InstallationManager/eclipse/
./IBMIM
- Create an ICM case, and verify BPM is working as expected by moving case one step.
/bpmdb/bpmdb/BAWscripts
Response File
Move the response.xml file to bpm app server review and update if any:
- Execute below command:
- After successful execution of above command, go in following directory and check files:
/bpmapp/bpmuser/ibm/BPM/v8.5/profiles/DmgrProfile/dbscripts/PSCell1.SBPPROD/DB2/CPEDB
- Copy these files to BPM database server, we need to execute them.
- ./createDatabase_ECM.sh
Review and update directory as per env
createTablespace_ECM_DOS.sql
- db2 -tvsf createTablespace_ECM_DOS.sql
Review and update directory as per env
createTablespace_ECM_TOS.sql
createTablespace_ICN.sql
Post Upgrade:
/bpmapp/bpmuser/ibm/BPM/v8.5/profiles/DmgrProfile/config/cells/PSCell1/nodes/Node1/
servers/AppClusterMember1/process-server/config
Changes to be made:
<bpd-engine>
<enable-legacy-behavior-for-dur-msg-when-msg-event-is-in-
loop merge="replace">true</enable-legacy-behavior-for-dur-
msg-when-msg-event-is-in-loop>
</bpd-engine>
Apply Fix-pack patch 8.6.20020001-WS-BPM-IFJR62453.zip Size = 247KB
3. Open Installation manager and Provide the zip file of fix pack in Add repository section of
Installation Manager. Click Update on main screen of installation manager.
1. 8.6.20020001-WS-BPM-IFJR62974.zip this fix pack was unistaled and applied new one by
IBM
2. 8.6.20020001-WS-BPM-IFJR62480.zip
Step 3: Open Installation manager and Provide the zip file of fix pack in Add repository section of
Installation Manager. Click Update on main screen of installation manager.
Click on upgrade
Once done with above step, please run below command (Path need to be changed
according to environment)
Open Installation manager and Provide the zip file of fix pack in Add repository section of Installation
Manager. Click Update on main screen of installation manager.
3.then open the first task (Register the Administrator Console for Content Engine (ACCE) Plug-in) and
run it
Open the case management dmgr proifle:
Register the Administrator Console for Content Engine (ACCE) Plug-in and run.
On successful completion. Restart bpm apps services.
After starting the server, Following commands should be executed on BPM application server:
AdminConfig.save()
exit