Opatchauto72030 Execute In Nonrolling Mode High Quality (PROVEN)
While it causes downtime, it is often faster for the overall maintenance window because patches are applied to the shared binaries once rather than node-by-node.
Review the log file located at $ORACLE_BASE/cfgtoollogs/opatchauto . Conclusion
Avoid running opatchauto against patch directories hosted on network file systems (NFS). Read/write latency or temporary network drops can corrupt the execution state, prompting a 72030 failure. Always stage patches on fast, local SSD/NVMe filesystems (e.g., /u01/stage ).
For high-availability systems, it is recommended to move workloads to a standby system (e.g., via Data Guard) before initiating a non-rolling patch session. Comparison: Rolling vs. Non-Rolling Rolling Mode (Default) Non-Rolling Mode ( -nonrolling ) Availability Continuous; services stay up on other nodes. Complete outage required; all nodes down. Compatibility Only for local/non-shared homes. Required for shared homes or non-rollable patches. Complexity Node-by-node sequential execution. Parallel/Simultaneous execution on shared binaries. opatchauto72030 execute in nonrolling mode high quality
Note: Ensure your OPatch utility is updated to the absolute latest version recommended by the specific quarterly RU readme. Step 2: Clear Pending Operating System Reboots
| Phase | Action Item | Completed? (Yes/No) | | :--- | :--- | :--- | | | Full backup of software, datafiles, and inventory. | ☐ | | Pre-Patching | Updated OPatch to the latest version (p6880880). | ☐ | | Pre-Patching | Verified SSH and sudo (if using non-root) is working. | ☐ | | Pre-Patching | Verified all nodes have sufficient disk space. | ☐ | | Pre-Patching | Ran opatch lsinventory on all nodes to confirm they are at the same level. | ☐ | | Pre-Patching | Ran opatchauto apply <patch> -analyze to detect conflicts. | ☐ | | Execution | Changed working directory from /root or / . | ☐ | | Execution | Ran opatchauto apply <patch> -nonrolling as root . | ☐ | | Post-Patching | Ran opatch lsinventory on all nodes to verify installation. | ☐ | | Post-Patching | Manually ran datapatch -verbose to apply SQL changes (if needed). | ☐ | | Post-Patching | Performed full cluster and database functionality tests. | ☐ |
The internal cluvfy or pre-patch checks failing to get a unanimous status from all RAC (Real Application Clusters) nodes. While it causes downtime, it is often faster
Confirm that the patch was applied correctly to both the Grid Infrastructure and Database homes:
# As grid user /u01/app/19.0.0/grid/OPatch/opatch lsinventory # As oracle user /u01/app/oracle/product/19.0.0/dbhome_1/OPatch/opatch lsinventory Use code with caution. Step 5: Restart Cluster Services and Databases
Major patch bundles containing infrastructure changes that prevent cross-node version mixing. Read/write latency or temporary network drops can corrupt
Run regular inventory verification checks ( opatch lsinventory -detail ) to ensure the central inventory remains uncorrupted and readable by the Oracle software installation owner.
OPatchAuto executes commands shifting between the root user and the Oracle software owner (e.g., oracle or grid ). If central inventory permissions ( oraInventory ) are misconfigured, or if SSH user equivalence between cluster nodes drops, the workflow breaks. 4. Locked Binaries or Stale Processes
To resolve the OPATCHAUTO-72030 error, you must explicitly tell opatchauto to use non-rolling mode by adding the -nonrolling flag. 1. Prerequisites Ensure all nodes are running current patches. Log in as the user. Prepare the patch directory. 2. Run the Patching Command
:
When running opatchauto apply in ( -nonrolling ), the utility requires absolute state consistency across nodes. The failure chain generally looks like this: