The pacemaker packages have been upgraded to upstream version 1.1.5, which provides a number of bug fixes over the previous version. (BZ#668466)

Due to missing Cluster Resource Manager (CRM) scripts, an attempt to run the crm utility caused the following message to be presented to a user:

crm_standby not available, check your installation

This update re-includes the "crm_master" in the packemaker package, and adapts the crm utility not to require the "crm_standby" and "crm_failcount" scripts, so that the above error message is no longer displayed. (BZ#627626)

In a cluster environment managed by both Pacemaker and the CMAN cluster management subsystem, frequent leaving and joining of a node could cause Pacemaker's quorum view to be incorrect. This update applies a patch that addresses this issue, so that the leaving and joining of a node no longer causes Pacemaker's quorum view to be different from CMAN's. (BZ#684825)

Previously, rebooting a node in a CMAN managed cluster could cause the fencing daemon to keep the "fenced:default" CPG group on the remaining nodes, leaving the cluster in an inconsistent state. With this update, an upstream patch has been applied to address this issue. As a result, when a node is rebooted and leaves a cluster, the cluster resources correctly run on remaining nodes. (BZ#684838)

When using the fence_ipmilan fencing agent, Pacemaker now accepts "diag" as a valid per-device stonith action. With this action enabled, a fenced node receives the "DIAG" signal and creates a dump for diagnostic purposes. (BZ#676286)

Note that Pacemaker provides the following functionality, and is now the preferred application to perform these tasks:

Pacemaker provides support for a time-based resource control. This allows system administrators to define the time a resource is down. (BZ#310361)

Pacemaker supports time-based resource control. This allows system administrators to define the downtime of individual resources. (BZ#449833)

Pacemaker allows users to manually start or stop cluster resources. This enables system administrators to perform maintenance tasks on individual components within a service. (BZ#449835)