Symptom
Extended downtime can occur when failover split-brain happens and then is resolved, because the primary unit delays its GARP until configuration is replicated to the secondary unit and is applied there.
The request is to send GARP earlier, as soon as split-brain is detected and resolved and secondary unit is forcefully moved to the cold standby state.
Conditions
All ASA versions and platforms are affected.
Workaround
Use debug menu ipaddrutl 5 command from the user context to send GARP manually for the IP address.
Further Problem Description
Split-brain shouldn't normally happen, but if it happens, user traffic can be blackholed for few more seconds or few more minutes after the split-brain condition is cleared, until config sync is complete, when both primary and secondary units send GARP.
