If the 72030 error occurred mid-patch and left the cluster in a partially patched state, you must resume the session without the offending flag. Clean up any dead lock files if prompted by the utility. Resume the orchestration: $GRID_HOME/OPatch/opatchauto resume Use code with caution.
Note: These are generic steps; always follow the specific patch README.
, your patching session has likely come to a grinding halt. This specific error typically triggers when you try to run opatchauto in the default rolling mode on a shared GI home configuration. Why This Happens By default, opatchauto apply
Before running in non-rolling mode, ensure the clusterware is ready to be stopped. Manually test the shutdown: # As root crsctl stop crs -f Use code with caution. opatchauto72030 execute in nonrolling mode exclusive
is a common sign that your environment requires a different patching strategy than the default. This error typically triggers when you attempt a rolling patch on a shared GI home Oracle documentation notes must be updated across all nodes simultaneously. What is OPATCHAUTO-72030? The error message explicitly states:
: Create an OS-level backup or snapshot of the shared Grid Infrastructure home and database homes.
When you run a command like the one below without the required parameters, you will be met with this error: If the 72030 error occurred mid-patch and left
Even after correcting the mode, you might encounter other issues. Here are common scenarios and solutions.
Run the command from the first node. The -exclusive flag ensures the GI stack is managed correctly during the downtime.
The term "exclusive" in this context usually refers to an or state required for patching shared binaries. Rocky Linux release 8.8 19c rac ru安装 - 墨天轮 Note: These are generic steps; always follow the
opatchauto patches one node at a time. The utility shuts down services on Node 1, applies the patch to local binaries, and brings the services back online. It then shifts to Node 2. The database remains accessible to users via surviving nodes throughout the entire process.
Would you like the exact syntax for opatchauto with -nonrolling and -exclusive for a given Oracle version?
The execute part of the patch may fail due to insufficient privileges or missing tablespaces.
By default, opatchauto applies patches in rolling mode unless you instruct it otherwise. When you run opatchauto without the -nonrolling flag, it defaults to rolling. However, a shared CRS home is incompatible with this process. The utility checks the environment, detects the shared home configuration, and correctly throws OPATCHAUTO-72030 to prevent an invalid operation.
Non-rolling mode is typically used when a patch affects the Oracle Clusterware itself, when patching a shared CRS home, or when the patch metadata mandates it. In a non-rolling patching session, all nodes (both local and remote) must be down or stopped before patching begins. The process involves patching the first node, then nodes 2 through n-1, and finally the last node. This mode results in full cluster downtime during the patching window.