The LC product had successfully installed onto the node, but failed to "ripple" across to the cell that was managed by a WebSphere Application Server 6.1 Network Deployment DM node. The error message being displayed was CLFRP1261E.
The lcinstalllog.txt wasn't particularly useful, in terms of providing further details.
As is always the way, I couldn't find owt about this message in the Information Centre or on't'internet etc. so had to rely upon native wit and intelligence ( OK, I guessed ).
I took a look at /tmp using the df -km command and - c'est voila - it was showing at 100% used.
I cleared /tmp down and re-ran the installation ( the second time around, one chooses to add a standalone installation to a cell/cluster ) and it just worked.
Therefore, I'm assuming that CLFRP1261E actually means "Hey dude, you've like got no frikking disk space in /tmp, man" or something.
Don't take my word for it, but it did work :-)
3 comments:
Yes, always check the temp drive/directory for space in addition to the overall disk space available.
Famous cause of Domino index failures and other odd items.
Excellent troubleshooting job!
Encountered the same error but different issue -
The following was reported in the install log (/tmp/lcinstalllog.txt)
[exec] ADMU0124I: The system clock of the new node (Feb 8, 2010 11:44:14 AM) is not synchronized with that of the deployment manager.
[exec] ADMU0125E: Change the clock of the new node to be within 5 minutes of the clock of the deployment manager.
There's a note in the InfoCenter stating that this sync should be ensured or done before clustering.
Once the clocks were sync'd, reran the install taking the option to convert a "stand-alone deployment" .. and was successful.
Hope this can be helpful some day ...
Otieno, definitely - syncing the clocks across the nodes within the cell/cluster is a mandatory requirement for any clustered environment, whether Connections, Portal or anything else, good tip, thanks
Post a Comment