Question

What are IBM BPM automated Backup best practices and standards?

Answer 1

Back up the four things that together make a BAW environment restorable, always at the same point in time:

1. Databases - BPMDB (Process Server / Center runtime and repository), PDWDB (performance data warehouse), CMNDB (cell / messaging), and with case management the CPE object stores (DOS / TOS / GCD databases) plus the FileNet file storage areas. Use the database's online backup with archived logs so that you can restore to a point in time; snapshot all BPM databases together (they reference each other by ids).
2. WebSphere profiles - manageprofiles -backupProfile for the deployment manager and every node, or a file-system backup of profiles/; this holds the cell configuration, 100Custom.xml, SSL keystores, installed applications.
3. Install root - one backup after every fix pack; plus the Installation Manager data directory (/var/ibm/InstallationManager) so that uninstall / update still work.
4. The design-time repository - export every process app / toolkit snapshot as .twx on a schedule (BPMExport command or the REST export) and keep them in source control; that is what you will actually use to rebuild a Process Center from scratch.

# nightly (Linux) - Db2 example
db2 backup db BPMDB online to /backup compress include logs
db2 backup db PDWDB online to /backup compress include logs
db2 backup db CMNDB online to /backup compress include logs
/opt/ibm/BAW/bin/manageprofiles.sh -backupProfile -profileName DmgrProfile -backupFile /backup/dmgr-$(date +%F).zip
/opt/ibm/BAW/bin/BPMExport.sh -name "Claims" -snapshot "1.4" -outputFile /backup/twx/Claims-1.4.twx -user admin -password ...

Standards that make the difference: test a restore quarterly into a separate cell; keep backups off the server (object storage); encrypt; keep the retention of BPM logs aligned with database point-in-time needs; stop the event manager (Process Admin > Event Manager > pause) before an offline backup so that no half-written instances exist; on CP4BA back up the persistent volumes and the databases the operator uses, and keep the custom resource YAML and secrets in Git.

References

• BAW documentation - backing up and restoring the environment