The health-check function is disabled on the switch by default. The function is automatically enabled when the first server member is added to the first SLB group on the switch, and it is disabled automatically when the last member is removed from the last SLB group on the switch.
Each SLB server member in an LBG is monitored by an ICMP ping originated from the switch. When ping fails, a failover event takes place. SLB supports an active-standby configuration. A standby server must be created in order for failover to take place. When a failure occurs in any of the active servers in the LBG, SLB looks for the first available standby server within the same SLB group to replace it. When a replacement server is found, all traffic hashed to the failed server will be redirected to the standby server. Traffic hashed to all other servers is not affected.