Lesson 4: Troubleshooting ZDM – Common Errors & Fixes
Ashish Agnihotri
Data Analyst at Exaguru
Introduction
Even though Oracle Zero Downtime Migration (ZDM) automates end-to-end migrations, DBAs often encounter environment-specific errors. These issues usually arise during pre-checks, staging, or cutover phases.
This lesson outlines the most common ZDM errors, their root causes, and fixes, following Oracle’s best practices.
🔹 ZDM Troubleshooting Architecture
ZDM logging and diagnostics are structured across three layers:
1. ZDM Service Host Logs
- Located Under:
$ZDM_HOME/logs/
- Key Files:
zdm.log → High-Level Job Log
zdmlogger.log → Component Logging Framework
job_<ID>.log → Task-level log per migration job
2. Source Database Logs
-RMAN logs, listener logs, alert logs (alert_<SID>.log).
3. Target Database Logs
-DBUA upgrade logs (for cross-version migrations).
- Standby creation logs (if Data Guard mode).
🔹 Common Errors & Fixes
1️⃣ Connectivity & Authentication Errors
Error Message:
ZDM-10030: Unable to establish SSH connectivity with source/target node
Cause:
- SSH key not configured correctly.
- Wrong hostnames or firewall restrictions.
Fix:
# Test SSH manually ssh -i /path/to/key opc@target-node
- Ensure passwordless SSH between ZDM host ↔ source ↔ target.
- Open required ports (22, 1521, Data Guard/GoldenGate ports).
2️⃣ Pre-check Failures
Error Message:
ZDM-10142: Database version not supported for migration
Cause:
- Source database not on terminal release (e.g., 11.2.0.4 required for 11g).
Fix:
- Patch/upgrade source to the minimum supported version.
- Example: 11.2.0.4 PSU applied before attempting migration.
3️⃣ RMAN Backup Errors
Error Message:
ORA-19504: failed to create file ORA-27040: file create error
Cause:
- Insufficient disk space for RMAN backup.
- Wrong staging location in ZDM config file.
Fix:
- Increase disk quota on ZDM host staging area.
- Validate
backupLocationparameter inzdm_service.cfg.
4️⃣ Data Guard Sync Issues
Error Message:
ORA-16810: multiple errors or warnings detected for the database
Cause:
- Data Guard broker misconfiguration.
- Network latency is impacting redo transport.
Fix:
DGMGRL> show configuration; DGMGRL> edit database <dbname> set property LogXptMode='SYNC';
- Validate redo apply lag.
- Ensure ports (default 1521) are open and the MTU is consistent.
5️⃣ GoldenGate Replication Errors
Error Message:
OGG-00519: Fatal error in Extract
Cause:
- Unsupported data type in replication.
- Extract not granted required privileges.
Fix:
GRANT DBA TO <ogg_user>;
- Add supplemental logging:
ALTER DATABASE ADD SUPPLEMENTAL LOG DATA;
- Restart GoldenGate extract/replicat.
6️⃣ Cutover Failures
Error Message:
ZDM-10502: Cutover failed due to application connections on source DB
Cause:
- Applications still connected during cutover.
Fix:
ALTER SYSTEM QUIESCE RESTRICTED;
- Stop application services.
- Retry cutover with:
zdmcli resume job -id <JOB_ID>
🔹 Best Practices for Troubleshooting
✅ Always run
-precheck before migration to catch environment mismatches. ✅ Monitor logs in real-time:
tail -f $ZDM_HOME/logs/job_<ID>.log
✅ Use
zdmcli query job to check job progress. ✅ Automate health checks with Ansible playbooks before ZDM runs.
✅ Document fixes in a runbook for repeatable migrations.
🔹 Example: Resume Failed Job
# Query the job status
zdmcli query job -id 12345
# Resume from failure
point zdmcli resume job -id 12345
🔹 Closing Thoughts
Troubleshooting ZDM requires understanding its multi-layered architecture: ZDM logs, source DB, target DB, and transport mechanisms (RMAN/DG/OGG). By mastering common error codes and their fixes, DBAs can ensure smooth migrations with minimal downtime and reduced risk.
This makes ZDM not only an automation tool but also a resilient migration framework when combined with strong troubleshooting practices.
Next Step: Lesson 5: Comparing ZDM vs GoldenGate vs RMAN
