08-18-2022 03:00 PM
We do node HA failover testing on a routine basis, currently we login to the grid and push buttons to force failover between nodes (pic below of what we choose, Grid - Grid Master - Members - checkmark the pair to HA - Control - Force HA Failover - Click Ok to confirm - Wait).
I am wondering if there is a wAPI/REST call to make this happen? I looked through the wAPI/REST guides and searched here but didnt find anything. I know sometimes there are hidden/undocumented APIs.
08-22-2022 09:38 AM
I looked through the WAPI documentation and I believe your assumption is correct, however, there may be a workaround. In the API call dhcpfailover, two function calls exist that may help:
set_dhcp_failover_partner_down: Force a HA member into a 'Partner Down' state
set_dhcp_failover_secondary_recovery: Force the failover recovery of the secondary member in the HA pair
These calls are on page 375 of the WAPI documentation (link). I cannot confirm that these calls will work the same as a 'Force HA failover' from the web interface, so I suggest testing these calls in a lab to confirm if they have the desired results.
08-23-2022 11:19 AM
HA failover (Failover between two members that are in the same cluster) is very differnt than a DHCP failover association (Failover beween two members or clusters). "set_dhcp_failover_partner_down" and "set_dhcp_failover_secondary_recovery" only affect the DHCP relationship, not between members of the same cluster.
08-29-2022 11:36 AM
The call you are looking for is a function call on the member object. Here is a generic link to the appropriate documentation:
Let me know if you need an example call.