- When moving services around a cluster, it was not possible to refer to virtual machines without the "vm:" prefix. With this update, the virtual machines can be controlled both with and without the "vm:" prefix.
- When a custom resource agent was configured without the primary attribute, the rgmanager utility terminated unexpectedly with a segmentation fault. The underlying source code has been modified to address this bug, and an error is logged instead of the crash of rgmanager in the described scenario.
- Previously, the rg_test utility did not handle exit codes correctly; when rg_test failed, the zero exit code was returned instead of a non-zero exit code. This update applies a patch to fix this bug and if rg_test crashes, the zero exit code is returned as expected.
- When the "time_t" values were not configured correctly in the cluster.conf file, the rgmanager could terminate unexpectedly. This update provides a patch to fix this bug, and rgmanager now properly handles the incorrectly configured values.
- When more instances were configured for a resource address than were specified with the "maxinstances" attribute, no message was logged. With this update, a warning is returned in such a case.
- When a resource start or status operation failed to execute due to resource starvation on the node, the user had to reboot the node manually. With this enhancement, a new reboot_on_pid_exhaustion flag has been added. The flag allows to set up rgmanager to reboot the node automatically in the described scenario.