Distributed Systems Administration Utilities V1.1 Release Notes for HP-UX 11i v2 March 2006

Configuration Synchronization
Problems that Have Been Fixed
• cfservd was not restarted after distributing new security keys to managed clients. This has been fixed.
• The “master;” line was not correctly written to the csync_wizard answer file when running the wizard
in a Serviceguard cluster. This has been fixed.
Known Problems and Workarounds
• When running the Configuration Synchronization Wizard in a Serviceguard cluster, make sure your
current directory is not on the filesystem associated with the shared storage for the package. When
creating the Serviceguard package, the filesystem will be forcibly dismounted and Serviceguard will
inadvertently kill the wizard process since it is preventing the filesystem from being dismounted.
• During each use, the Configuration Synchronization Wizard creates backups of the current cfengine
configuration in the directory /var/opt/dsau/cfengine/backups. The top level files in this directory are
the most recent backups. Any previous configurations are saved time-stamped subdirectories named
"v_timestamp”. These backup directories are never deleted and can accumulate if the wizard is used
many times.
Command Fanout
Parallel Distributed Shell (pdsh), Parallel Distributed Copy (pdcp) and the commands based on these utilities
(cexec, ccp, ckill, cuptime, cps) display misleading error messages when the security environment is not
properly configured. For example, if non-interactive ssh is not properly configured, pdsh and pdcp display
error messages such as the following:
pdsh@source_node:target_node:Failed to initiate RCP protocol
pdsh@source_node:target_node:ssh exited with exit code 255
This indicates that pdsh was unable to establish an ssh connection with the remote node. For additional
troubleshooting information, refer to Managing Systems and Workgroups , Chapter 3, Configuring a System.
6 System Administration Release Notes