8.198. resource-agents

Bug Fixes

BZ#1025909

Previously, the IPaddr2 agent did not send out unsolicited neighbor advertisements to announce a link-layer address change. Consequently, floating IPv6 addresses, which require this functionality, could not work correctly. To fix this bug, the "send_ua" internal binary required for IPaddr2 agent to drive IPv6 addresses has been added. As a result, the floating IPv6 addresses now work correctly. Nevertheless, IPv4 addresses are left unaffected by this change.

BZ#1025504

Previously, when an oracledb.sh resource was created without the "listener_name" option, the rgmanager oracledb resource agent failed to start. With this update, the oracledb.sh file has been modified to operate correctly when no "listener_name" is given, and thus oracledb now starts correctly.

BZ#1024065

Prior to this update, the netfs agent could become unresponsive during a stop operation, even with the "self_fence" option enabled. With this update, the self fence operation is executed sooner in the process, which ensures that NFS client detects server leaving if umount can not succeed, and self fencing occurs.

BZ#1054327

Previously, the virtual machine (VM) instances managed by the VirtualDomain agent failed during the monitor operation if the libvirtd daemon was not available. This caused the Pacemaker resource manager to unnecessarily recover healthy VM resources when libvirtd failed. With this update, the VirtualDomain agent is capable of monitoring KVM virtual machines without requiring libvirtd to be accessible. As a result, Pacemaker no longer mismanages VM resources in case of a libvirtd failure.

BZ#993329

Previously, the "no_unmount" functionality was missing in the netfs.sh file. Consequently, the netfs resource agent did not allow an NFS share to remain mounted after a service was relocated. This update adds back the missing functionality, and the file system is now left mounted when the service relocates.

BZ#1022793

The following agents were shipped in error, and thus have now been dropped: nginx, rsyslog, mysql-proxy, and slapd.

BZ#1023340

Previously, the SAPInstance resource agent for Pacemaker did not behave correctly on nodes where the corosync node name did not match the host name. The check provided by the sapinstance_notify() function has been modified to fix this bug. The SAPInstance agent now works correctly regardless of the match between the corosync node name and host name.

BZ#1091101

Previously, Pacemaker's nfsserver resource agent was unable to properly perform NFSv3 network status monitor (NSM) state notifications. As a consequence, NFSv3 clients could not reclaim file locks after server relocation or recovery. This update introduces the nfsnotify resource agent, thanks to which NSM notifications can be sent correctly, thus allowing NFSv3 clients to reclaim file locks.