Procedure to update new license in running cluster...
1.] Put the cluster into a safe mode by freezing all Service Groups. This is necessary to avoid unwanted Service Group failover during the maintenance time.
The command can be executed from any one node for each Service Group is:
# hagrp -freeze <group name> -persistent
2.] Create a backup folder for the current installed licenses on each cluster node-
# cp -pR /etc/vx/licenses/lic /etc/vx/licenses/lic_bkp
3.] Remove all licenses on each cluster node -
# ls -l /etc/vx/licenses/lic/
# rm -rf /etc/vx/licenses/lic/*
# ls -l /etc/vx/licenses/lic
4.] Install a new license on each node
Note - One license key at a time and repeat this process to install multiple licenses.
# /opt/VRTS/bin/vxlicinst -k <License Key>
5.] To update licensing information in a running Volume manager on each cluster node
# vxdctl license init
== With the init argument specified, the vxdctl license init command requests that vxconfigd reread any persistently stored license information. If licenses have expired, this can make some features unavailable. If new licenses were added, but are not yet recognized by vxconfigd, this command forces vxconfigd to rescan the licenses, and to make the new licenses available.
# vxdctl mode
== The mode should be enabled.
== This command should show all available features.
== This command should show expected output.
NOTE- The step 1-5 must be run on each node in the cluster before executing the next set of commands.
6.] To update new licenses in a running cluster -
a.) Update system-level licensing information on all nodes in the cluster. Execute below step on all nodes in the cluster -
# hasys -updatelic -all
b.) Update cluster-level licensing information from any node in the cluster -
# haclus -updatelic
# hastatus -sum
The above commands should show the expected output of the cluster.
If the licensing logging still occurs even after applying the first method it means that the Veritas Cluster Server CmdServer process needs to be restarted to get the new license mode. To do that please run the commands below:
1. Stop the CmdServer process on all the nodes within the cluster
# /opt/VRTSvcs/bin/CmdServer -stop
2. Restart the CmdServer on all the nodes
7.] Unfreeze all Service Groups in the cluster-
Please use the command on all (previously frozen) Service Groups:
# hagrp -unfreeze <group name> -persistent