VCS MultiNICB Agent does not work properly with Solaris 10

The MultiNICB resource goes to FAULTED  state, without any clear reason

Error Message

VCS WARNING V-16-10001-6519 MultiNICB:.....:monitor:Deleting default router (a8fa9fa)
VCS INFO V-16-1-10307 Resource (Owner: unknown, Group: ....) is offline on .. (Not initiated by VCS)


Cause

Starting with Solaris 10 the path of in.mpathd command was changed: if MultiNICB resources were configured to work with previous Solaris release, they will not loger work properly.

The default path was:
/sbin/in.mpathd
this is now a symbolic link to:
/usr/lib/inet/in.mpathd

so must be changed in resource configuration otherwise MultiNICB Agent will not be able to determine if process is running and kill/restart when requested


Solution

Change the path of mpathd command with in MultiNICB  resource:

# haconf -makerw
modify the path:
# hares -modify <MultiNICB_res> MpathdCommand "/usr/lib/inet/in.mpathd -a"
# haconf -dump -makero
then check if the new attribute is properly set:
#  hares -display <MultiNICB_res> -attribute MpathdCommand
#Resource               Attribute                         System   Value
<MultiNICB_res>    MpathdCommand          global     /usr/lib/inet/in.mpathd -a

Restart the Agent on all of cluster nodes

# haagent -stop MultiNICB -sys <node_name>
# haagent -start MultiNICB -sys <node_name>