Pip: If you've ever watched a file system clone grind to a halt at ninety percent and thought "well, at least it's almost done," Syed Anwar Ahmed has a post that will gently correct your optimism.
Mara: This episode covers one of those Oracle EBS R12.2 maintenance failures where the error message on the surface points one direction and the real problem is buried several log files deeper — specifically inside WebLogic credential configuration during a patch file system clone.
Pip: Let's start with what actually went wrong during that fs_clone operation, and why the obvious error wasn't the guilty party.
—
Pip: If you've ever watched a file system clone grind to a halt at ninety percent and thought "well, at least it's almost done," Syed Anwar Ahmed has a post that will gently correct your optimism.
Mara: This episode covers one of those Oracle EBS R12.2 maintenance failures where the error message on the surface points one direction and the real problem is buried several log files deeper — specifically inside WebLogic credential configuration during a patch file system clone.
Pip: Let's start with what actually went wrong during that fs_clone operation, and why the obvious error wasn't the guilty party.
When the Clone Lies to Your Face
Pip: The setup here is a routine EBS R12.2 maintenance cycle — adop phase=fs_clone is supposed to rebuild the patch file system, and instead it keeps failing. The surface error looks like an edition problem, which sends you hunting in entirely the wrong direction.
Mara: Right, and the post is explicit about this misdirection. The initial message the operation reported was "ERROR: Unable to set edition. Invalid edition type : PATCH," followed by "adclone.pl did not go through successfully. EXIT STATUS: 1."
Pip: So you see "Invalid edition type : PATCH" and you start thinking about edition configuration — which is exactly the wrong rabbit hole, because the actual failure is a WebLogic password decryption error sitting deeper in the logs.
Mara: The WLST logs showed the real exception: the failure chain ran from txkSetAppsLibsTarget.py through setNodeManagerCredentials and getNodeManagerPassword, landing on "com.rsa.jsafe.JSAFE_PaddingException: Invalid padding." WebLogic could not decrypt the Node Manager password while configuring fs2.
Pip: What this means in practice is that the clone stopped because of a credential encryption mismatch — not disk space, not port conflicts, not anything the OS was doing. The post confirms all three of those were ruled out: filesystem space was adequate, ports were acquired cleanly, and OS logs showed no OOM kills or Java process terminations.
Mara: The ninety percent progress point on the FMW Home cloning step was a location marker, not a cause. The post is careful to say that figure shows where the clone stopped, not why.
Pip: And the fix, once the actual cause was identified, is a five-step sequence: correct boot.properties, restart WebLogic, run AutoConfig on the RUN file system, run Preclone on the RUN file system, then rerun fs_clone with force=yes.
Mara: The post closes with a pointed takeaway — messages like "adclone.pl did not go through successfully" only tell you the clone failed. The actionable diagnosis requires going into the FSCloneApplyAppsTier and WLST logs to find the component that actually broke. The JSAFE_PaddingException pattern specifically signals a WebLogic credential and encryption configuration issue, not a resource problem.
Pip: Worth noting: the post flags that this resolution is environment-specific. JSAFE_PaddingException during EBS cloning can have different underlying causes, so reviewing the detailed logs before applying this fix elsewhere is the actual lesson — not just copying the five steps.
Mara: Exactly the kind of maintenance incident where the log discipline matters as much as the fix itself.
Pip: The real takeaway is that generic failure messages are just the beginning of the investigation, not the answer.
Mara: Dig into the WebLogic and WLST logs — that's where the actual story lives. More from the blog next time.
Leave a comment