Sorry, I do not think that executing "clusterXL_admin down -p" will help you in this case.
First of all the "-p" flag is for "permanent", meaning it is intended use is to survive reboot. To simply cause a failover, the
"clusterXL_admin down" should be executed on the Active member. Attention!!! please verify that your cluster is in a good working state before doing it.
Additionally, if your VSX cluster is running in VSLS mode, performing "clusterXL_admin down" in VS0 will not cause the other VS' running on Active member to failover:sk95133
In case of VSLS, use vsx_util to redistribute VS' to the desired target VSX unit before executing "clusterXL_admin down" in VS0.
Regards,
Vladimir