IPMULTINICB Resource Faults with the error as IPMultinicB:monitor:Could not read MultiNICB export information.

Messages from /var/VRTSvcs/log/engine_A.log:

VCS NOTICE V-16-10001-5001 (lfs8606q)IPMultiNICB:lfs8606q_IPMultinicB:monitor:Could
not read MultiNICB export information. This may be because MultiNICB has not completed monitor yet.

 

 
Possible Causes:

1) Incorrect types.cf file.
2) Incorrect permissions on the export file /var/VRTSvcs/lock/MultiNICB/<MultiNICBResName>
3) Non-existent export file /var/VRTSvcs/lock/MultiNICB/<MultiNICBResName>.
4) The child MultinicB resource is not probed prior to the parent IPMultinicB resource.


Solution:

It appears that the files /var/VRTSvcs/lock/MultiNICB/<MultiNICBResName>_ MultinicB_export are not being generated by the MultiNICB agent.

Import the correct types.cf file from /etc/VRTSvcs/conf directory to /etc/VRTSvcs/conf/config/ directory as follows.

Stop VCS without Affecting Service Groups . (Execute on 1 node only)

 

#hastop -all -force

Take a Backup of existing types.cf as follows..  (Execute on1 node)

#mv /etc/VRTSvcs/conf/config/types.cf /etc/VRTSvcs/conf/config/types.cf.incorrect

Copy over the actual types.cf file as follows..  (Execute on the same node as above)

#cp /etc/VRTSvcs/conf/types.cf /etc/VRTSvcs/conf/config/types.cf

Start Veritas Cluster server Engine on all the nodes..
#hastart   (on the node first where you have copied the types.cf file).


It may be beneficial to allow MultiNICB more time to write the status of the interfaces to the export file by increasing the IPMultiNICB's' MonitorInterval attribute to 60 ( default 30 ).


# haconf -makerw
# hatype -modify IPMultiNICB MonitorInterval 60
# haconf -dump -makero